Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Characterising Jupiter's dynamo radius using its magnetic energy spectrum

Characterising Jupiter's dynamo radius using its magnetic energy spectrum Jupiter’s magnetic field is generated by the convection of liquid metallic hydrogen in its interior. The transition from molecular hydrogen to metallic hydrogen as temperature and pressure increase is believed to be a smooth one. As a result, the electrical conductivity in Jupiter varies continuously from being negligible at the surface to a large value in the deeper region. Thus, unlike the Earth where the upper boundary of the dynamo—the dynamo radius—is definitively located at the core-mantle boundary, it is not clear at what depth dynamo action becomes significant in Jupiter. In this paper, using a numerical model of the Jovian dynamo, we examine the magnetic energy spectrum at di erent depth and identify a dynamo radius below which (and away from the deep inner core) the shape of the magnetic energy spectrum becomes invariant. We find that this shift in the behaviour of the magnetic energy spectrum signifies a change in the dynamics of the system as electric current becomes important. Traditionally, a characteristic radius derived from the Lowes–Mauersberger spectrum—the Lowes radius—gives a good estimate to the Earth’s core-mantle boundary. We argue that in our model, the Lowes radius provides a lower bound to the dynamo radius. We also compare the Lowes–Mauersberger spectrum in our model to that obtained from recent Juno observations. The Lowes radius derived from the Juno data is significantly lower than that obtained from our models. The existence of a stably stratified region in the neighbourhood of the transition zone might provide an explanation of this result. Keywords: Jupiter; dynamo region; anelastic convection; magnetic energy spectrum; Lowes–Mauersberger spectrum 1. Introduction As existing technology does not allow us to take direct mea- surement inside Jupiter, or for that matter, inside the Earth, we Jupiter has the strongest magnetic field among the planets have to deduce the dynamo radius of a planet from measure- in the Solar System. The magnitude of its surface magnetic ments made near its surface. In the case of the Earth, where the field is about ten times larger than that of the Earth. Jupiter and dynamo radius is at the core-mantle boundary, Lowes (1974) the Earth both have dipole-dominated magnetic fields, with the introduced a strategy by considering the average magnetic en- dipolar axis inclined at about 10 to the rotation axis. However, ergy over a spherical surface of radius r, the recent NASA Juno mission (Bolton et al., 2017) revealed 1 1 that Jupiter’s magnetic field has its non-dipolar part mostly con- E (r) = jB(r; ; )j sin  d d: (1) fined to the northern hemisphere (Moore et al., 2018; Jones, 2 4 2018), unlike the Earth’s field which shows no such preference. Here B is the magnetic field, (r; ; ) are the standard spherical The intricate magnetic field of Jupiter is believed to be gener- coordinates based on the rotation axis,  is the permeability of ated by the convective stirring of liquid metallic hydrogen in free space and the time argument t has been suppressed. From the planet’s interior. An important and long-standing question the bottom of the insulating mantle up to the planetary surface, is at what depth does such dynamo action begin. The dynamo there is no electric current j = 0. The magnetic field in this radius—the location of the top of the dynamo region—is an im- region can thus be written as B = rV . The scalar potential V portant factor in understanding the interaction between the in- satisfies r V = 0 and is given by, terior magnetohydrodynamics and the atmospheric flow in the outer layer (Cao and Stevenson, 2017). Knowledge of the dy- 1 l X X l+1 namo radius also provides constraints that help to improve the a V (r; ; ) = a P (cos ) estimation of electrical properties inside Jupiter and will sub- l=1 m=0 sequently lead to better modelling of the Jovian magnetic field. m m (g cos m + h sin m) (2) l l The dynamo radius also determines where internal torsional os- cillations in Jupiter are reflected as they propagate outwards where a is a reference radius often taken to be the mean plan- (Hori et al., 2019). etary radius. P are the Schmidt semi-normalised associated m m Legendre polynomials. The Gauss coecients g and h are l l determined from magnetic field measurement at the surface. Corresponding author Email address: y.tsang@leeds.ac.uk (Yue-Kin Tsang) Preprint submitted to Earth and Planetary Science Letters December 20, 2019 arXiv:1905.07661v2 [astro-ph.EP] 19 Dec 2019 Figure 1: (a) Normalised Lowes spectrum R =R at r = r calculated from the Gauss coecients in the JRM09 model of Connerney et al. (2018). A linear fit to l 1 J log R (r ) for 2  l  10 gives the Lowes radius r = 0:845r . Changing the fitting range to 5  l  10 results in r = 0:796r . Connerney et al. (2018) 10 l J lowes J lowes J suggest the data is compatible with r = 0:87r . Here, r = 6:9894 10 m. (b) The non-zonal part of the JRM09 data is compared with the full spectrum. Both lowes J J spectra are normalised by the value of R of the full spectrum. Note that the non-zonal data gives a much closer fit to a straight line in the range 6  l  10 than the full data and a linear fit in this range gives r = 0:828r . (c) Comparison of normalised Lowes spectrum R =R at r = r from the Juno data JRM09 and our lowes J l 1 J Jupiter dynamo model at Pm = 10 and Pm = 3. Using the expression (2) in (1) yields reasonably with seismic measurement. Langlais et al. (2014) found r = 3294:5 km compared to the seismically deter- lowes mined 3481:7 km. Langlais et al. (2014) also found that omit- 2 E (r) = R (r) (3) 0 B l ting the m = 0 axisymmetric components in (4), so that only l=1 the non-zonal components are used, where 2l+4 h i nz m 2 m 2 2l+4 X R (r) = (l + 1) (g ) + (h ) ; (8) h i a l l l m 2 m 2 r R (r) = (l + 1) (g ) + (h ) (4) l m=1 l l m=0 reduced the scatter of the spectrum and led to a remarkably ac- is the Lowes spectrum, or sometimes the Lowes–Mauersberger curate agreement between r and the Earth’s seismic core lowes spectrum (Mauersberger, 1956). It follows that radius. Compared to the Earth, magnetic field measurements for 2l+4 Jupiter are less extensive. The data available before the Juno R (r) = R (a): (5) l l mission only allowed for the calculation of the Lowes spec- trum up to l = 4 (Connerney et al., 1998) or l = 7 (Ridley The downward continuation relation (5) gives the Lowes spec- and Holme, 2016) depending on the modelling methodology. trum R (r) at some depth r in terms of R (a) at the surface. It l l This has changed since the Juno spacecraft arrived at Jupiter. relies crucially on B being purely potential. While the spacecraft is taking more measurements as it contin- To estimate the depth of the dynamo region, we need one ues to orbit Jupiter, Connerney et al. (2018) computed g and further assumption. It has been argued that the large-scale part h up to l = 10 from the data collected during eight of the first of R (a) mainly originates from the Earth’s outer core and turbu- nine flybys. Using these Gauss coecients, Fig. 1(a) shows the lence there results in a uniform distribution of magnetic energy Lowes spectrum at r = r where, following French et al. (2012), over di erent scales l. In particular, at some depth r near lowes we take the core-mantle boundary, R (r ) is independent of l. This l lowes r = 6:9894 10 m (9) ‘white source hypothesis’ (Backus et al., 1996), together with (5) implies the linear relation to be the mean radius of Jupiter. Note that Connerney et al. (2018) used the equatorial radius, 7:1492 10 m, and we have log R (a)  (a)l (6) corrected for this di erence in Fig. 1 and Table 1. In sharp con- trast to its Earth counterpart (Backus et al., 1996), the Jovian for the large scales with (a) satisfying Lowes spectrum R (r ) does not show a clean exponential de- l J (a)=2 cay. In fact R (r ) remains almost constant for 2  l  5 before l J r = 10  a: (7) lowes decaying at larger l. Consequently, routinely applying Lowes’ Thus, (7) gives the Lowes radius r in terms of the spectral procedure gives di erent values of r depending on the range lowes lowes slope which can be determined solely from magnetic mea- of l used in the linear fit, as shown in Fig. 1(a). However, using surement at the surface. The Lowes radius provides an estimate the non-zonal components only on the JRM09 data, see equa- to the location of the Earth’s core-mantle boundary that agrees tion (8), as suggested by Langlais et al. (2014) for the Earth, 2 leads to a better linear fit for l  6, see Fig. 1(b). The best-fit value of r = 0:828r . This improvement may arise because lowes J the higher order non-axisymmetric field components arise more directly from the non-axisymmetric convection and hence are more randomly distributed than the full spectrum components. A more fundamental issue here concerns the interpreta- tion of r for Jupiter. The interior structure of Jupiter is lowes very di erent from that of the Earth. Theoretical and experi- mental studies suggest that the phase transition from molecu- lar to degenerate metallic hydrogen (Wigner and Huntington, 1935) along a Jupiter adiabat is continuous (Wicht et al., 2018; Helled, 2018). As a result, the electrical conductivity (r) of the hydrogen-helium mixture in Jupiter varies smoothly with the radial distance (Weir et al., 1996; Liu et al., 2008). For ex- ample, Fig. 2 shows the profile (r) obtained from an ab initio simulation by French et al. (2012). Therefore unlike the Earth, where the flow and the Lorentz force acting on the flow are both confined within the same region, Jupiter’s dynamo is cou- Figure 2: Electrical conductivity as a function of depth r along a Jupiter adia- pled to an outer layer of fluid flow that is free from magnetic bat. The circles are the J11-8a model from the ab initio simulation of French et al. (2012). The solid line is the hyperbolic model (14). The values of the pa- e ects. Such coupling, with a transition layer in between, is not 6 8 rameters are c = 4:279 10 , c = 274:9, c = 2:544 10 , c = 1:801 1 2 3 4 well understood. It is not clear how large the current-free region and c = 20:28. The inset plots the same data in linear scale. where the downward continuation operation (5) is justified ac- tually is. There is also the question about the validity of the white source hypothesis. In fact, how do we characterise the 2.1. Anelastic spherical dynamo extent of the dynamo region for a continuously varying elec- We consider the convection of an electrically conducting trical conductivity profile? Is there a sensible way to define fluid in a rotating spherical shell of inner radius r = 0:092r in J a dynamo radius for Jupiter? In this paper, we examine these and outer radius r = 0:959r . The heat flux is modelled out J issues by considering the magnetic energy spectrum F (r), to by an entropy flux proportional to the local entropy gradient be defined in section 3, in a numerical model of Jupiter. The with constant di usivity  . The other physical parameters are magnetic energy spectrum F (r) essentially represents the dis- the angular speed , the constant kinematic viscosity  and tribution of magnetic energy over di erent spherical harmonic the magnetic di usivity (r) which varies with the radial dis- degrees l at depth r. The Lowes spectrum R (r) in (4) is a special tance r. The dynamical variables of velocity, magnetic field, case of F (r) under the condition j = 0 (which is only true near entropy, density and pressure (u; B; S; ; p) are governed by the the planetary surface). The change in behaviour of F (r) along non-dimensional equations, r indicates varying dynamics in di erent regions. Comparing F (r) to R (r) gives further insights into the di erent physics in Ek Du EkRaPm dT l l 0 0 + 2zˆ  u = r S rˆ these regions. Pm Dt Pr dr (10a) In the next section, we describe our model for Jupiter’s dy- 1 F + (r B) B + Ek ; namo. In section 3, we first introduce the magnetic energy spec- ¯  ¯ trum and discuss how its behaviour changes with depth. We @B then show that a dynamo radius can be identified from a transi- = r (u B)r (r B); (10b) @t tion in the spectral slope. In section 4, we look at the relation- DS Pm ship between the dynamo radius and the Lowes radius. We then ¯ ¯ ¯T = r ( ¯TrS ) Dt Pr finish with a discussion on the di erences between results from (10c) Pm Pr 1 our model and observation. +  ¯T H + Q + Q ; S  J Pr RaPm Ek r ( ¯u) = 0; (10d) 2. A model of Jupiter’s dynamo r B = 0; (10e) Numerical models of Jupiter’s dynamo have recently been developed to study the magnetic field and internal flow of the gi- together with the equation of state ant planet (Jones et al., 2011; Jones, 2014; Gastine et al., 2014). The model used in the present study is developed and described @ @  ¯ dT 1 d ¯ d = dS + d p = dS d p: (11) in detail by Jones (2014), though here the range of parameters @S @ p g ¯ dr  ¯g ¯ dr p S has been extended to get further into the strong-field dynamo In deriving the model, the Lantz and Fan (1999) formulation regime (Dormy, 2016). We briefly summarise it here. of the anelastic approximation (Braginsky and Roberts, 2007) has been employed about a spherically symmetric, hydrostatic 3 ¯ and adiabatic basic state (; ¯ p ¯; S ). This simplifies the system to involve the dynamics of a single thermodynamical variable S = S S . In (10a), zˆ and rˆ are unit vectors along the rotation 0 0 0 axis and the radial direction respectively,  = p = ¯ +  is a generalised pressure combining disturbance pressure p = p p ¯ and disturbance gravitational potential  . In (11), g ¯ is the gravitational acceleration due to  ¯. The system is forced by a constant entropy source H modelling the secular cooling of the planet. The dissipative terms are: @u @ @u 2 @ j i F =  ¯ + ( ¯r u); (12a) ;i @x @x @x 3 @x j i j i j=1 X 2 1 @u @u 2 Q =  ¯ + (r u) ; (12b) 2 @x @x 3 i j i; j=1 Q = jr Bj : (12c) Let T ,  and  be the dimensional values of T ,  ¯ and , re- spectively, at the midpoint of the shell and S be the entropy drop across the thickness L = r r of the shell. Our equa- out in tions are non-dimensionalised using the unit of length L, time L = and magnetic field , where  is the perme- 0  0 ability of free space. The dimensionless numbers in (10) are defined as: Figure 3: Simulation at Pm = 10. (a) Dipole-dominated radial magnetic field T L S at r = r . (b) Radial magnetic field at r = 0:8r has larger magnitude and out J Ra = ; Ek = ; Pr = ; Pm = : (13) more small-scale structures. (c) Left: zonal velocity u at r = r showing the S S   out prograde equatorial jet. Right: zonal and time averaged u . The boundary condition for u is no-slip at r and stress-free at in r . At both the inner and outer boundaries, it is electrically out insulating and S is fixed at a constant value. The initial con- 2.3. Simulation parameters and electric current profile ditions are u = 0 with a small perturbation in B and S . The For the rest of this paper, the following parameters are kept spectrum of the initial magnetic perturbation is narrow-banded 7 5 fixed: r =r = 0:0963, Ra = 2  10 , Ek = 1:5  10 and in out with 8  l  10 and thus has no dipole component. Pr = 0:1. Jupiter is believed to have a strong-field dynamo, i.e., its magnetic field strongly influences the flow. Thus, we are in- 2.2. Hyperbolic electrical conductivity profile terested in cases of large magnetic Prandtl number Pm which In applying the anelastic convective system described above produce a strong-field dynamo (Dormy, 2016). Specifically, to model the Jovian dynamo, we need to provide an equilib- we investigate the e ects of Pm by comparing simulations with rium state and a conductivity profile that represent the thermo- Pm = 10 and Pm = 3. The case of Pm = 3 has previously been dynamic and transport properties inside Jupiter. French et al. studied in detail by Jones (2014). Results here are presented in (2012) have calculated the material properties of a hydrogen- dimensionless units unless otherwise stated. helium-water mixture under Jupiter-like condition using den- Figure 3 shows snapshots of magnetic and velocity fields sity functional theory. Here, we use the same equilibrium den- from the Pm = 10 simulation. The equatorial zonal jet and the sity  ¯(r) and temperature T (r) profile as in Jones (2014) which dipolar nature of the radial magnetic field, both near the surface, are smooth interpolations to the J11-8a data in French et al. are obvious. With the conductivity (r) increasing sharply with (2012). For the magnetic di usivity (r), we consider the fol- depth, we are interested in the radial dependence of the dynamo lowing hyperbolic model: action. We first examine the average electric current j at a rms (ln  + c r + c )(ln  + c r + c ) = c (14) 1 2 3 4 5 depth r, defined as, with five parameters. The values used for these parameters are D E 2 2 j (r) = j j(r; ; ; t)j sin dd: (15) the same as in Jones (2014) which give a good fit to the J11-8a rms data. Figure 2 shows the corresponding electrical conductivity (r) = 1= . In the interior, (r) decreases roughly linearly Above, j = r  B, hi indicates time average over a statisti- and reaches about one-fifth of its maximum value at r = 0:8r . cal steady state and the integral is over a spherical surface of This is unlike some of the previous studies (Gastine et al., 2014; radius r. Figure 4 shows that j (r) behaves very similarly for rms Glatzmaier, 2018) in which the electrical conductivity is taken both Pm = 3 and Pm = 10. In the interior where strong mag- to be constant below a certain depth. Dietrich and Jones (2018) netic field is being generated, j (r) increases slightly with r rms studied a wide range of profiles for (r) by varying c in (14) and peaks at around r = 0:7r even though (r) is monotoni- 2 J and found a diversity of magnetic field morphologies. cally decreasing in r. Approaching the surface, j (r) follows rms 4 Figure 5: Time evolution of selected modes of the magnetic energy spectrum F (r ; t) at the outer boundary r = r for the case of Pm = 10. The dipolar l out out l = 1 mode (dashed line) eventually becomes dominant as the system reaches a statistical steady state at about t = 0:05. Table 1 For simulations at two di erent Pm and Juno observation (with the two fitting ranges in Fig. 1(a)): (r ) is the spectral slope of F (= R at r ) measured at Figure 4: Average current j , defined in (15), at di erent depth r for Pm = 10 out l l out rms the outer boundary r and is measured inside the upper dynamo region and Pm = 3. The electrical conductivity (r) from Fig. 2 is also plotted on the out dyn as discussed below (21). r is the Lowes radius in (7) and r is the dy- right axis. lowes dyn namo radius in (22). Rm(r) is the depth-dependent magnetic Reynolds number defined in (25) and r is where Rm = 1. r = 6:9894 10 m. Rm J the trend of (r) and drops quickly and smoothly to negligi- (r ) r =r r =r Rm(r ) r =r out dyn lowes J dyn J dyn Rm J ble value, indicating the cessation of dynamo action. While the Pm = 10 0.072 0.024 0.883 0.907 156 0.939 variation of j (r) with r certainly signifies di erent dynamics Pm = 3 0.089 0.035 0.865 0.900 111 0.934 rms Juno (l  2) 0.109 ? 0.845 ? ? ? at di erent depth, Fig. 4 does not locate a characteristic depth Juno (l  5) 0.162 ? 0.796 ? ? ? that represents the top of the dynamo region. It is not obvious from the profile of j (r) at what depth the electric current be- rms comes large enough to generate a significant magnetic field. In interpreted as the average magnetic energy per spherical har- the next section, we show that a dynamo radius can be identified monic degree l (Maus, 2008). Note that F is calculated from using the magnetic energy spectrum. the full field B. On the other hand, R is calculated from the scalar potential in (2). Nevertheless, in a current-free region, 3. Magnetic energy spectrum the two are identical. Hence, We again consider the average magnetic energy on a spher- F (r) = R (r) if j (r) = 0: (19) l l rms ical surface given by (1). Unlike in the derivation of the Lowes spectrum R where the magnetic field is assumed to be potential, l 3.1. F (r) at the planetary surface here we make no such assumption and expand B in terms of a We first examine F (r) at the outer boundary r = r of l out m m m set of vector spherical harmonics fY ; ; g (see Appendix l l l the spherical shell. Figure 5 shows the time evolution of a few A), modes of F (r ; t) from the Pm = 10 simulation. The devel- l out 1 l X X m m m opment of the dipolar l = 1 mode can clearly be seen as it B = q Y + s + t  : (16) lm lm lm l l l outgrows all other modes. l=1 m=l Since the electric current is negligible at r , F (r ) = out l out The expansion coecients are generally function of r and t. R (r ) in our simulations, see for example Fig. 6 for the case l out Substituting (16) into (1), we get of Pm = 10. Figure 1(c) plots F from the Pm = 10 and Pm = 3 simulations as well as R calculated from the Juno data JRM09. All three spectra have been continued to r = r using (5) and 2 E (r; t) = F (r; t) (17) 0 B l J l=1 normalised. The key qualitative di erence between the spec- tra from our two simulations is that F (r ) for Pm = 3 displays l J where the magnetic energy spectrum is a clear exponential decay for all l (excluding l = 1) while for Pm = 10, the spectrum is roughly flat at small l and only starts 4 3 m;0 2 2 2 to decay exponentially for l & 5. In this respect, the Pm = 10 F (r; t) = jq j +js j +jt j : (18) l lm lm lm (2l + 1) spectrum is similar to the Juno spectrum. However, the decay m=0 rate of F for Pm = 3 is faster and slightly closer to the Juno We are mainly interested in the time-averaged spectrum F (r) = observed value. Fitting the range 5  l  40 yields the values hF (r; t)i , with the time-averaging done after the system has l t of r shown in Table 1 for Pm = 10 and Pm = 3. lowes reached a statistical stationary state. Roughly, F (r) can be 5 Figure 7: Spectral slope (r) of the magnetic energy spectrum F (r) and the spectral slope (r) of the Lowes spectrum R (r) in the Pm = 10 simulation. At the dynamo radius r , diverges from and remains more or less constant dyn about inside a large part of the dynamo region. The Lowes radius r is dyn lowes Figure 6: Time-averaged magnetic energy spectrum F (r) (solid lines), defined where the downward continued R is flat, (r ) = 0. l lowes in and below (17), and Lowes spectrum R (r) (circles) at di erent depth r for the Pm = 10 simulation. R (r) are obtained from the time-averaged R (r ) l l out using (5). from R (r). The main observation in Fig. 6 is that the shape of F (r) becomes independent of r, as discussed in the previous section, at essentially the same depth r . This implies that dyn 3.2. F (r) at di erent depth r r is the boundary below which electric current becomes im- dyn We now look at the magnetic energy spectrum in the inte- portant and the dynamics of the system is altered. We therefore rior of the spherical shell. We focus on the case of Pm = 10. identify r as the top of the dynamo region, or the dynamo dyn The solid lines in Fig. 6 show F (r) at di erent depth r. The key radius. feature is the transition of F (r) through three di erent stages Figure 7 vividly illustrates the discussion in the previous as r varies. We have seen in the previous section the steep ex- paragraph by plotting (r) together with the Lowes spectral ponential decay of F with l (for l & 5) at the surface. As we slope (r), defined analogously to (6), as a function of r. From move below the surface, Fig. 6 shows F (r) maintains such ex- (5), we have ponential decay, out (r) = (r ) 2 log ; (21) out log F (r)  (r)l; (20) where (r ) = (r ). Note how (r) diverges from (r) and out out levels o to the value at r = r . The sharpness of the dyn dyn but the spectral slope (r) decreases rapidly with r inside the transition allows for a meaningful definition of r . Figure 7 layer of 0:91r . r  r and F (r) has become rather shal- dyn J out l also provides a quantitative way to determine r . Fitting a dyn low at 0:91r . As we delve further into the interior, quite re- horizontal line through 0:7r < r < 0:86r yields the value of J J markably, F (r) remains shallow as its shape, and hence (r), = 0:024. We then obtain the dynamo radius r from (21) dyn dyn becomes more or less invariant over the substantial region of using the relation 0:55r . r . 0:91r . This clearly indicates a shift in the dy- J J (r ) = : (22) dyn dyn namics near 0:91r . Finally, in the deep interior and close to the core, F (r) decays super-exponentially and the magnetic field This gives r = 0:907r for the Pm = 10 simulation. The dyn J is dominated by large scales. Boussinesq geodynamo models values of the various spectral slopes and characteristic radii for which compute the magnetic energy spectrum inside the core Pm = 10 are summarised in Table 1. also show a spectrum that decays exponentially with l for l  5, Spacecraft missions can only measure the spectral slope at e.g. Christensen et al. (1999). the planetary surface, from which r is calculated using the lowes white source assumption discussed in section 1. The question 3.3. A dynamo radius is then how well can r predict the actual dynamo radius lowes The significance of the change in behaviour of F (r) be- r . In our simulations, r is where the dashed line in Fig. 7 dyn lowes comes clear when we compare F (r) to the Lowes spectrum intersects the horizontal axis, i.e. (r ) = 0 in (21). This is lowes R (r) at the same depth. Recall that R (r ) = F (r ) at the because by definition, the downward continued R (r) becomes l l out l out surface. We now downward continue R (r ) using (5) (with flat at r = r , at least for the range of l where R (r ) is fitted l out lowes l out a = r ) to obtain R (r) at di erent r, which we plot as circles to obtain (r ). It is clear from Fig. 7 that generally out l out in Fig. 6. Near the surface, F (r) and R (r) are essentially in- l l r  r : (23) distinguishable because electric current is negligible there. As lowes dyn r decreases and reaches some depth r , F (r) starts to deviate dyn l 6 3.4. E ects of Pm We now compare the simulation at Pm = 10 to the one at Pm = 3. We have already discussed the di erences in the magnetic energy spectrum at the surface in section 3.1. The slightly steeper F (r ) for Pm = 3 means the magnetic field l out has less small scales. Generally, results from the two simula- tions are qualitatively similar. The Pm = 10 simulation has stronger magnetic fields, so the Elsasser number, which is pro- portional to the root-mean-squared magnetic field over the do- main, is about an order of magnitude larger than that of the Pm = 3 simulation, closer to the high value expected in Jupiter. In the Pm = 3 simulation, it was noted in Jones (2014) that the Lorentz force in the interior mostly suppresses the inter- nal di erential rotation in the metallic hydrogen region. At Pm = 10 the stronger magnetic field means the internal di er- ential rotation is even more strongly suppressed, see Fig. 3(c), so that strong zonal flow occurs only in the molecular region near the equator. However, although the zonal flow is confined to the molecular region, its magnitude is about double that of the Pm = 3 simulation. The convective velocity in the Pm = 10 simulation is also about double that in the Pm = 3 simulation, so the magnetic Reynolds number is about twice that of the Pm = 3 simulation. While it is not computationally possible to achieve the parameters believed to operate in Jupiter, the in- crease in Pm has moved the simulation results in the direction of more realistic values. Figure 4 shows the electric current is roughly three times smaller in the Pm = 3 case. The spectral slopes for Pm = 3 Figure 8: Selected modes of the magnetic energy spectrum F (r) and the Lowes display the same trend as in Fig. 7. The values of (r ) and out spectrum R (r) as a function of depth r in the simulation at Pm = 10. (a) F (r) l l over the whole spherical shell. (b) F (r) (solid lines) and R (r) (dotted lines) together with that of r and r are given in Table 1. l l dyn lowes dyn near a region about the dynamo radius r . dyn A steeper spectrum at the surface means a bigger (r ). At out the same time, we see that also increases. The net result dyn is that the dynamo radius r is only marginally less than that dyn Comparing the values shown in Table 1, we see that for Pm = of Pm = 10. On the other hand, r drops more significantly lowes 10, r is about 3% less than r . The di erence stems from lowes dyn which makes r less accurate as a predictor of r . These lowes dyn being fairly small but not exactly zero. Comparing F to dyn l results suggest that for a dynamo with a larger Pm, the magnetic R near r in Fig. 6 again shows the white source assump- l lowes energy spectrum F (r) in the upper part of the dynamo region tion is only approximate and F (r) never becomes exactly flat. is closer to being ‘white’. As a consequence, the Lowes radius Nonetheless, helped by the steep decrease of (r) with r shown gives a better prediction to the dynamo radius. in Fig. 7, we still have a close agreement between r and lowes r . Figure 3(b) shows a snapshot of the radial magnetic field dyn at r = 0:8r where (0:8r )  . 4. Discussion and conclusion J J dyn Figure 8 reveals further details about the radial dependence The electrical conductivity in Jupiter varies from being neg- of the magnetic energy spectrum by plotting a selected number ligible at the surface to a very high value in the interior. It of modes of F (r) as a function of r. The first few modes (l . 5) thus raises the question about the depth at which dynamo action are exceptional as they varies irregularly with r while the rest starts. In this paper, we consider the magnetic energy spectrum of the modes are well represented by l = 20 in Fig. 8(a). The F (r) at depth r in a numerical model of Jupiter’s dynamo. For l = 1 mode has the largest magnitude at all r except for a region l & 5, the magnetic energy spectrum decays exponentially with around 0:8r where it is overtaken by l = 2, see also Fig. 3(b). l, log F (r)  (r)l. We find that a sharp transition in (r) Interestingly, this is also the region in which j (r) peaks and 10 rms can be used to identify a dynamo radius r and this dynamo dyn (r) remains virtually constant at . We also see that all dyn radius can be reasonably predicted by the Lowes radius r lowes non-dipolar modes are strongly damped just above the dynamo as discussed in section 3.3. The situation is in fact rather sim- radius r leading to the dipole-dominated field observed at dyn ple as illustrated in Fig. 7. The two characteristic radii r and dyn the surface. Figure 8(b) zooms into a region about r = r and dyn r are controlled by two spectral slopes: (r ) which is ob- lowes out shows, for the large-scale modes, how F (r) deviates from R (r) l l servable at the surface and which measures the deviation dyn as one moves from the current-free outer layer into the dynamo from the white source hypothesis near the top of the dynamo region. 7 However, Rm does give a somewhat smaller r  0:930r r Rm J and Rm (r )  O(10) for Pm = 10. The three magnetic r dyn Reynolds numbers in (25) all use the shell thickness L as the typical length scale. An alternative is to use the magnetic dif- fusivity scale height d (r) = (d=dr) . Dietrich and Jones (2018) and Wicht et al. (2019) have examined, among other things, the characterisation of Jovian dynamo models using dif- ferent magnetic Reynolds numbers. Since d  L, we expect Rm  Rm  O(10) and Rm  O(1) at r in our simulations r dyn if the definition of Rm based on d (r) is used. On the other hand, numerical dynamo models cannot achieve the large value of Rm expected in the interior of Jupiter. If we adopt the velocity of 2 1 U = 10 m s which has been estimated for Jupiter’s interior, see e.g. Jones (2014), and use our values of L and , then the value of r at which Rm = U L= = 1 increases to r = 0:957r . We have estimated the Lowes radius from the Juno data in Fig. 1. We find that it has fairly large uncertainties depending on the range of l used in the linear fitting. However, if the zonal components are omitted we find a closer linear fit for l  6 as did Langlais et al. (2014) for the geomagnetic data. The situ- ation in Jupiter could be similar to our simulation at Pm = 10 Figure 9: Depth-dependent magnetic Reynolds numbers, defined in (25), based where a clean exponential decay in the spectrum emerges only on di erent velocity scales for Pm = 10 and Pm = 3. The two vertical dot- dashed lines indicate the locations of r and r given in Table 1. dyn lowes at larger l. This is in contrast to the case of Pm = 3, which has weaker flow and magnetic field. We anticipate further data col- lected by the ongoing and future flybys will extend the range of region. Varying moves the horizontal dot-dashed line in dyn the Lowes spectrum in Fig. 1 and hence provide a more reliable Fig. 7 up and down while changing (r ) shifts the dashed out estimate of r . lowes curve left and right. These determine the location of r and dyn Despite uncertainties in the data, Table 1 shows that r lowes r as well as the relative distance between them. Notice that lowes from the Juno observation is clearly smaller than in both of our the dashed curve, given by (21), is essentially a straight line for simulations. Nevertheless, its implication on the location of the r r  r , out out dynamo radius is not clear. Whether r gives a good esti- lowes 2r mate on r relies on the white source hypothesis, which may dyn (r)  2 + (r ): (24) out or may not be valid in Jupiter. However, the Pm = 10 simula- out tion, which we believe is closer to Jupiter conditions than the We find that in our two simulations at Pm = 10 and Pm = 3, Pm = 3 simulation, has a smaller than the Pm = 3 sim- dyn (r ) and change in such a way that leaves the dynamo out dyn ulation, suggesting that Jupiter’s magnetic field might be close radius fairly insensitive to Pm. Incidentally, at r = r , the dyn to white near the top of the dynamo region. In our simulations, electrical conductivity (r) has dropped by two orders of mag- a steeper spectrum observed at the surface tends to be accom- nitude from its maximum at the inner boundary. panied by steeper spectra in the interior and consequently r dyn Figure 9 plots as a function of depth three di erent magnetic could be shallower than r by a fair amount. The present lowes Reynolds numbers, each based on a di erent velocity scale, results suggest that the true dynamo radius likely lies above the Lu (r) Lu (r) Lu (r) Lowes radius. ;rms rms r;rms Rm = ; Rm = ; Rm = ; Irrespective of its relation to the dynamo radius, the Lowes (r) (r) (r) (25) radius is a property of the magnetic field at the surface of for our two Jovian dynamo simulations, reverting to dimen- Jupiter. The smaller r of the Juno data stems from a steeper lowes sional units. Here, u , u and u are the root-mean- spectrum at the planetary surface, implying Jupiter’s magnetic rms r;rms ;rms squared values of the total, radial and zonal velocity, respec- field has less small scales than that in our model. This is slightly tively, over a spherical surface of radius r. All three magnetic surprising as the flow is believed to be more vigorous in Jupiter Reynolds numbers vary weakly in the interior and then decrease than in our simulations, because the simulations have enhanced sharply to a negligible value near the surface. We also see that di usion coecients to maintain numerical stability. We should Rm  Rm for r & r indicating the zonal flow becomes point out again that r for the simulations in Table 1 are dyn lowes dominant. This is consistent with the depth of the equatorial derived from time-averaged spectra while the Juno observa- zonal jet estimated roughly from figures such as Fig. 3. The tion essentially provides only a snapshot of Jupiter’s magnetic values of Rm at r and the depth r at which Rm = 1 are field. Figure 10 plots the instantaneous Lowes radius r (t) dyn Rm lowes given in Table 1. These values suggest that r determined obtained from the time-dependent spectrum F (r ; t) in the Rm l out by the criterion Rm = 1 is generally larger than r estimated Pm = 10 simulation. The case of Pm = 3 shows similar spread dyn from F (r). Using Rm instead of Rm gives similar estimates. about the mean value. We argue that the di erence between 8 Acknowledgements The authors are supported by the Science and Tech- nology Facilities Council (STFC), ‘A Consolidated Grant in Astrophysical Fluids’ (grant numbers ST/K000853/1 and ST/S00047X/1). This work used the DiRAC@Durham facility managed by the Institute for Computational Cosmology on be- half of the STFC DiRAC HPC Facility (www.dirac.ac.uk). The equipment was funded by BEIS capital funding via STFC cap- ital grants ST/P002293/1, ST/R002371/1 and ST/S002502/1, Durham University and STFC operations grant ST/R000832/1. DiRAC is part of the National e-Infrastructure. Figure 10: Instantaneous Lowes radius r (t) obtained from the time- lowes dependent magnetic energy spectrum F (r ; t) at the outer boundary for Pm = l out 10. The dashed line is the time-averaged value over the statistical steady state Appendix A. Vector spherical harmonics given in Table 1. Following Barrera et al. (1985) but using the Schmidt semi- normalised associated Legendre polynomials P in the defini- simulations and observation is significant even when statistical tion of the spherical harmonics, fluctuation is taken into account. jmj The di erence in r between our numerical model and m im lowes Y (; ) = P (cos )e ; (A.1) l l observation raises several questions and suggests possible av- enues for future research. The results presented here are spe- we define three vector spherical harmonics: cific to the electrical conductivity profile (r) in (14) which m m is based on data from theoretical ab initio calculation. Could Y (; ) = Y rˆ; (A.2a) l l the deeper Lowes radius in observation mean the actual electri- m m (; ) = p rrY ; (A.2b) cal conductivity inside Jupiter is smaller than predicted? It is l l l(l + 1) worth studying how the magnetic energy spectrum responds to m m (; ) = rˆ  ; (A.2c) l l perturbations in (r). On a related note, the magnetic Reynolds number Rm in our simulations is of the order 10 at its max- which form an orthogonal basis for all square-integrable vector imum, much lower than an estimated value of 10 in Jupiter fields on the unit sphere. The (semi-)normalisation condition is (Jones, 2014). The puzzle here is that increasing Rm in the model will likely increase r rather than reduce it and thus lowes 0 4 m m 0 0 Y  (Y 0 ) sin dd = (2  )  ; (A.3) m;0 ll mm l l move it further away from the Juno value. While current com- 2l + 1 puting resources prevent us from reaching a much larger Rm, m m with similar expressions for and  . investigating the trend of the dynamics in the neighbourhood l l of a smaller attainable Rm, possibly by changing (r), could provide valuable insights. References In our simulations, the system is forced by the constant en- Backus, G., Parker, R., Constable, C., 1996. Foundations of Geomagnetism. tropy source H in (10c) and we employ a constant entropy Cambridge University Press, Cambridge. boundary condition. Could our numerical setup tend to pro- Barrera, R. G., Estev ´ ez, G. A., Giraldo, J., 1985. Vector spherical harmonics duce extra small scales that lead to the shallower spectra shown and their application to magnetostatics. Eur. J. Phys. 6, 287–294. in Fig. 1(c)? In geodynamo simulations, boundary conditions Bolton, S., Levin, S., Bagenal, F., 2017. Juno’s first glimpse of Jupiter’s com- plexity. Geophys. Res. Lett. 44, 7663–7667. can significantly a ect the dynamics (Sakuraba and Roberts, Braginsky, S. I., Roberts, P. H., 2007. Anelastic and Boussinesq approxima- 2009; Dharmaraj and Stanley, 2012). It is important to assess tions. In: Gubbins, D., Herrero-Bervera, E. (Eds.), Encyclopedia of Geo- the robustness of the present results and examine their depen- magnetism and Paleomagnetism. Springer Netherlands, pp. 11–19. dence on boundary conditions and the form of forcing. Cao, H., Stevenson, D. J., 2017. Zonal flow magnetic field interaction in the semi-conducting region of giant planets. Icarus 296, 59–72. The formation of a stably stratified layer just under the Christensen, U., Olson, P., Glatzmaier, G. A., 1999. Numerical modelling of the molecular layer due to ‘helium rain’ (Stevenson and Salpeter, geodynamo: a systematic parameter study. Geophys. J. Int. 138, 393–409. 1977) has been proposed to explain the near-axisymmetric Connerney, J. E. P., Acuan, ˜ M. H., Ness, N. F., Satoh, T., 1998. New models of magnetic field of Saturn (Stevenson, 1982; Dougherty et al., Jupiter’s magnetic field constrained by the Io flux tube footprint. J. Geophys. Res. 103, 11929–11939. 2018). Although helium rain is more probable to occur in Sat- Connerney, J. E. P., Kotsiaros, S., Oliversen, R. J., Espley, J. R., Joergensen, urn, it cannot be ruled out for Jupiter (Wahl et al., 2017; Debras J. L., Joergensen, P. S., Merayo, J. M. G., Herceg, M., Bloxham, J., Moore, and Chabrier, 2019). It would be interesting to see the e ects K. M., Bolton, S. J., Levin, S. M., 2018. A new model of Jupiter’s magnetic field from Juno’s first nine orbits. Geophys. Res. Lett. 45, 2590–2596. of such a stable layer on the Lowes radius as it displaces the dy- Debras, F., Chabrier, G., 2019. New models of Jupiter in the context of Juno namo action deeper into the interior. This is perhaps the most and Galileo. Astrophys. J. 100, 872. natural way to explain the surprisingly low value of r in the lowes Dharmaraj, G., Stanley, S., 2012. E ect of inner core conductivity on planetary Juno data. dynamo models. Physics of the Earth and Planetary Interiors 212-213, 1–9. 9 Dietrich, W., Jones, C. A., 2018. Anelastic spherical dynamos with radially variable electrical conductivity. Icarus 305, 15–32. Dormy, E., 2016. Strong-field spherical dynamos. J. Fluid Mech. 789, 500–513. Dougherty, M. K., Cao, H., Khurana, K. K., Hunt, G. J., Provan, G., Kellock, S., Burton, M. E., Burk, T. A., Bunce, E. J., Cowley, S. W. H., Kivelson, M. G., Russell, C. T., Southwood, D. J., 2018. Saturn’s magnetic field revealed by the Cassini Grand Finale. Science 362 (6410), eaat5434. French, M., Becker, A., Lorenzen, W., Nettelmann, N., Bethkenhagen, M., Wicht, J., Redmer, R., 2012. Ab initio simulations for material properties along the Jupiter adiabat. Astrophys. J. Suppl. 202, 5 (11pp). Gastine, T., Wicht, J., Durate, L. D. V., Heimpel, M., Becker, A., 2014. Ex- plaining Jupiter’s magnetic field and equatorial jet dynamics. Geophys. Res. Lett. 41, 5410–5419. Glatzmaier, G. A., 2018. Computer simulations of Jupiter’s deep internal dy- namics help interpret what Juno sees. PNAS 115, 6896–6904. Helled, R., 2018. The interiors of Jupiter and Saturn. Oxford Research Ency- clopedias: Planetary Science. Hori, K., Teed, R. J., Jones, C. A., 2019. Anelastic torsional oscillations in Jupiter’s metallic hydrogen region. Earth Planet. Sci. Lett., accepted. Jones, C. A., 2014. A dynamo model of Jupiter’s magnetic field. Icarus 241, 148–159. Jones, C. A., 2018. Jupiter’s magnetic field revealed. Nature 561, 36–37. Jones, C. A., Boronski, P., Brun, A. S., Glatzmaier, G. A., Gastine, T., Miesch, M. S., Wicht, J., 2011. Anelastic convection-driven dynamo benchmarks. Icarus 216, 120–135. Langlais, B., Amit, H., Larnier, H., Thebault, ´ E., Mocquet, A., 2014. A new model for the (geo)magnetic power spectrum, with application to planetary dynamo radii. Earth Planet. Sci. Lett. 401, 347–358. Lantz, S. R., Fan, Y., 1999. Anelastic magnetohydrodynamic equations for modeling solar and stellar convection zones. Astrophys. J. Suppl. 121, 247– Liu, J., Goldreich, P. M., Stevenson, D. J., 2008. Constraints on the deep-seated zonal winds inside Jupiter and Saturn. Icarus 196, 653–664. Lowes, F. J., 1974. Spatial power spectrum of the main geomagnetic field and extrapolation to the core. Geophys. J. R. Astron. Soc. 36, 717–730. Mauersberger, P., 1956. Das mittel der energiedichte des geomagnetischen ¨ ¨ ¨ hauptfeldes an der erdoberflache und seine sakulare anderung. Gerlands Beitr. Geophys 65, 207–215. Maus, S., 2008. The geomagnetic power spectrum. Geophys J. Int. 174, 135– Moore, K. M., Yadav, R. K., Kulowski, L., Cao, H., Bloxham, J., Connerney, J. E. P., Kotsiaros, S., Jørgensen, J. L., Merayo, J. M. G., Stevenson, D. J., Bolton, S. J., Levin, S. M., 2018. A complex dynamo inferred from the hemispheric dichotomy of Jupiter’s magnetic field. Nature 561, 76–78. Ridley, V. A., Holme, R., 2016. Modeling the Jovian magnetic field and its secular variation using all available magnetic field observations. J. Geophys. Res. Planets 121, 309–337. Sakuraba, A., Roberts, P. H., 2009. Generation of a strong magnetic field using uniform heat flux at the surface of the core. Nature Geoscience 2, 802–805. Stevenson, D. J., 1982. Reducing the non-axisymmetry of a planetary dynamo and an application to Saturn. Geophys. Astrophys. Fluid Dyn. 21, 113–127. Stevenson, D. J., Salpeter, E. E., 1977. The dynamics and helium distribution in hydrogen-helium fluid planets. Astrophys. J. Suppl. 35, 239–261. Wahl, S. M., Hubbard, W. B., Militzer, B., Guillot, T., Miguel, Y., Movshovitz, N., Kaspi, Y., Helled, R., Reese, D., Galanti, E., Levin, S., Connerney, J. E., Bolton, S. J., 2017. Comparing Jupiter interior structure models to Juno gravity measurements and the role of a dilute core. Geophys. Res. Lett. 44, 4649–4659. Weir, S. Y., Mitchell, A. C., Nellis, W., 1996. Metallization of fluid molecular hydrogen at 140GPa (1.4Mbar). Phys. Rev. Lett. 76, 1860–1863. Wicht, J., French, M., Stellmach, S., Nettelmann, N., Gastine, T., Durate, L., Redmer, R., 2018. Modeling the interior dynamics of gas planets. In: Luhr ¨ , H., Wicht, J., Gilder, S. A., Holschneider, M. (Eds.), Magnetic Fields in the Solar System. Vol. 448 of Astrophysics and Space Science Library. Springer Cham, Ch. 2, pp. 7–81. Wicht, J., Gastine, T., Durate, D. V., 2019. Dynamo action in the steeply de- caying conductivity region of Jupiter-like dynamo models. J. Geophys. Res. Planets 124, 837–863. Wigner, E., Huntington, H. B., 1935. On the possibility of a metallic modifica- tion of hydrogen. J. Chem. Phys 3, 764–770. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Astrophysics arXiv (Cornell University)

Characterising Jupiter's dynamo radius using its magnetic energy spectrum

Tsang, Yue-Kin; Jones, Chris A.
Astrophysics , Volume 2019 (1905) – May 18, 2019
Download PDF
10 pages

Loading next page...
 
/lp/arxiv-cornell-university/characterising-jupiter-s-dynamo-radius-using-its-magnetic-energy-qPCwCOwp4F

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

ISSN
0012-821X
eISSN
ARCH-3330
DOI
10.1016/j.epsl.2019.115879
Publisher site
See Article on Publisher Site

Abstract

Jupiter’s magnetic field is generated by the convection of liquid metallic hydrogen in its interior. The transition from molecular hydrogen to metallic hydrogen as temperature and pressure increase is believed to be a smooth one. As a result, the electrical conductivity in Jupiter varies continuously from being negligible at the surface to a large value in the deeper region. Thus, unlike the Earth where the upper boundary of the dynamo—the dynamo radius—is definitively located at the core-mantle boundary, it is not clear at what depth dynamo action becomes significant in Jupiter. In this paper, using a numerical model of the Jovian dynamo, we examine the magnetic energy spectrum at di erent depth and identify a dynamo radius below which (and away from the deep inner core) the shape of the magnetic energy spectrum becomes invariant. We find that this shift in the behaviour of the magnetic energy spectrum signifies a change in the dynamics of the system as electric current becomes important. Traditionally, a characteristic radius derived from the Lowes–Mauersberger spectrum—the Lowes radius—gives a good estimate to the Earth’s core-mantle boundary. We argue that in our model, the Lowes radius provides a lower bound to the dynamo radius. We also compare the Lowes–Mauersberger spectrum in our model to that obtained from recent Juno observations. The Lowes radius derived from the Juno data is significantly lower than that obtained from our models. The existence of a stably stratified region in the neighbourhood of the transition zone might provide an explanation of this result. Keywords: Jupiter; dynamo region; anelastic convection; magnetic energy spectrum; Lowes–Mauersberger spectrum 1. Introduction As existing technology does not allow us to take direct mea- surement inside Jupiter, or for that matter, inside the Earth, we Jupiter has the strongest magnetic field among the planets have to deduce the dynamo radius of a planet from measure- in the Solar System. The magnitude of its surface magnetic ments made near its surface. In the case of the Earth, where the field is about ten times larger than that of the Earth. Jupiter and dynamo radius is at the core-mantle boundary, Lowes (1974) the Earth both have dipole-dominated magnetic fields, with the introduced a strategy by considering the average magnetic en- dipolar axis inclined at about 10 to the rotation axis. However, ergy over a spherical surface of radius r, the recent NASA Juno mission (Bolton et al., 2017) revealed 1 1 that Jupiter’s magnetic field has its non-dipolar part mostly con- E (r) = jB(r; ; )j sin  d d: (1) fined to the northern hemisphere (Moore et al., 2018; Jones, 2 4 2018), unlike the Earth’s field which shows no such preference. Here B is the magnetic field, (r; ; ) are the standard spherical The intricate magnetic field of Jupiter is believed to be gener- coordinates based on the rotation axis,  is the permeability of ated by the convective stirring of liquid metallic hydrogen in free space and the time argument t has been suppressed. From the planet’s interior. An important and long-standing question the bottom of the insulating mantle up to the planetary surface, is at what depth does such dynamo action begin. The dynamo there is no electric current j = 0. The magnetic field in this radius—the location of the top of the dynamo region—is an im- region can thus be written as B = rV . The scalar potential V portant factor in understanding the interaction between the in- satisfies r V = 0 and is given by, terior magnetohydrodynamics and the atmospheric flow in the outer layer (Cao and Stevenson, 2017). Knowledge of the dy- 1 l X X l+1 namo radius also provides constraints that help to improve the a V (r; ; ) = a P (cos ) estimation of electrical properties inside Jupiter and will sub- l=1 m=0 sequently lead to better modelling of the Jovian magnetic field. m m (g cos m + h sin m) (2) l l The dynamo radius also determines where internal torsional os- cillations in Jupiter are reflected as they propagate outwards where a is a reference radius often taken to be the mean plan- (Hori et al., 2019). etary radius. P are the Schmidt semi-normalised associated m m Legendre polynomials. The Gauss coecients g and h are l l determined from magnetic field measurement at the surface. Corresponding author Email address: y.tsang@leeds.ac.uk (Yue-Kin Tsang) Preprint submitted to Earth and Planetary Science Letters December 20, 2019 arXiv:1905.07661v2 [astro-ph.EP] 19 Dec 2019 Figure 1: (a) Normalised Lowes spectrum R =R at r = r calculated from the Gauss coecients in the JRM09 model of Connerney et al. (2018). A linear fit to l 1 J log R (r ) for 2  l  10 gives the Lowes radius r = 0:845r . Changing the fitting range to 5  l  10 results in r = 0:796r . Connerney et al. (2018) 10 l J lowes J lowes J suggest the data is compatible with r = 0:87r . Here, r = 6:9894 10 m. (b) The non-zonal part of the JRM09 data is compared with the full spectrum. Both lowes J J spectra are normalised by the value of R of the full spectrum. Note that the non-zonal data gives a much closer fit to a straight line in the range 6  l  10 than the full data and a linear fit in this range gives r = 0:828r . (c) Comparison of normalised Lowes spectrum R =R at r = r from the Juno data JRM09 and our lowes J l 1 J Jupiter dynamo model at Pm = 10 and Pm = 3. Using the expression (2) in (1) yields reasonably with seismic measurement. Langlais et al. (2014) found r = 3294:5 km compared to the seismically deter- lowes mined 3481:7 km. Langlais et al. (2014) also found that omit- 2 E (r) = R (r) (3) 0 B l ting the m = 0 axisymmetric components in (4), so that only l=1 the non-zonal components are used, where 2l+4 h i nz m 2 m 2 2l+4 X R (r) = (l + 1) (g ) + (h ) ; (8) h i a l l l m 2 m 2 r R (r) = (l + 1) (g ) + (h ) (4) l m=1 l l m=0 reduced the scatter of the spectrum and led to a remarkably ac- is the Lowes spectrum, or sometimes the Lowes–Mauersberger curate agreement between r and the Earth’s seismic core lowes spectrum (Mauersberger, 1956). It follows that radius. Compared to the Earth, magnetic field measurements for 2l+4 Jupiter are less extensive. The data available before the Juno R (r) = R (a): (5) l l mission only allowed for the calculation of the Lowes spec- trum up to l = 4 (Connerney et al., 1998) or l = 7 (Ridley The downward continuation relation (5) gives the Lowes spec- and Holme, 2016) depending on the modelling methodology. trum R (r) at some depth r in terms of R (a) at the surface. It l l This has changed since the Juno spacecraft arrived at Jupiter. relies crucially on B being purely potential. While the spacecraft is taking more measurements as it contin- To estimate the depth of the dynamo region, we need one ues to orbit Jupiter, Connerney et al. (2018) computed g and further assumption. It has been argued that the large-scale part h up to l = 10 from the data collected during eight of the first of R (a) mainly originates from the Earth’s outer core and turbu- nine flybys. Using these Gauss coecients, Fig. 1(a) shows the lence there results in a uniform distribution of magnetic energy Lowes spectrum at r = r where, following French et al. (2012), over di erent scales l. In particular, at some depth r near lowes we take the core-mantle boundary, R (r ) is independent of l. This l lowes r = 6:9894 10 m (9) ‘white source hypothesis’ (Backus et al., 1996), together with (5) implies the linear relation to be the mean radius of Jupiter. Note that Connerney et al. (2018) used the equatorial radius, 7:1492 10 m, and we have log R (a)  (a)l (6) corrected for this di erence in Fig. 1 and Table 1. In sharp con- trast to its Earth counterpart (Backus et al., 1996), the Jovian for the large scales with (a) satisfying Lowes spectrum R (r ) does not show a clean exponential de- l J (a)=2 cay. In fact R (r ) remains almost constant for 2  l  5 before l J r = 10  a: (7) lowes decaying at larger l. Consequently, routinely applying Lowes’ Thus, (7) gives the Lowes radius r in terms of the spectral procedure gives di erent values of r depending on the range lowes lowes slope which can be determined solely from magnetic mea- of l used in the linear fit, as shown in Fig. 1(a). However, using surement at the surface. The Lowes radius provides an estimate the non-zonal components only on the JRM09 data, see equa- to the location of the Earth’s core-mantle boundary that agrees tion (8), as suggested by Langlais et al. (2014) for the Earth, 2 leads to a better linear fit for l  6, see Fig. 1(b). The best-fit value of r = 0:828r . This improvement may arise because lowes J the higher order non-axisymmetric field components arise more directly from the non-axisymmetric convection and hence are more randomly distributed than the full spectrum components. A more fundamental issue here concerns the interpreta- tion of r for Jupiter. The interior structure of Jupiter is lowes very di erent from that of the Earth. Theoretical and experi- mental studies suggest that the phase transition from molecu- lar to degenerate metallic hydrogen (Wigner and Huntington, 1935) along a Jupiter adiabat is continuous (Wicht et al., 2018; Helled, 2018). As a result, the electrical conductivity (r) of the hydrogen-helium mixture in Jupiter varies smoothly with the radial distance (Weir et al., 1996; Liu et al., 2008). For ex- ample, Fig. 2 shows the profile (r) obtained from an ab initio simulation by French et al. (2012). Therefore unlike the Earth, where the flow and the Lorentz force acting on the flow are both confined within the same region, Jupiter’s dynamo is cou- Figure 2: Electrical conductivity as a function of depth r along a Jupiter adia- pled to an outer layer of fluid flow that is free from magnetic bat. The circles are the J11-8a model from the ab initio simulation of French et al. (2012). The solid line is the hyperbolic model (14). The values of the pa- e ects. Such coupling, with a transition layer in between, is not 6 8 rameters are c = 4:279 10 , c = 274:9, c = 2:544 10 , c = 1:801 1 2 3 4 well understood. It is not clear how large the current-free region and c = 20:28. The inset plots the same data in linear scale. where the downward continuation operation (5) is justified ac- tually is. There is also the question about the validity of the white source hypothesis. In fact, how do we characterise the 2.1. Anelastic spherical dynamo extent of the dynamo region for a continuously varying elec- We consider the convection of an electrically conducting trical conductivity profile? Is there a sensible way to define fluid in a rotating spherical shell of inner radius r = 0:092r in J a dynamo radius for Jupiter? In this paper, we examine these and outer radius r = 0:959r . The heat flux is modelled out J issues by considering the magnetic energy spectrum F (r), to by an entropy flux proportional to the local entropy gradient be defined in section 3, in a numerical model of Jupiter. The with constant di usivity  . The other physical parameters are magnetic energy spectrum F (r) essentially represents the dis- the angular speed , the constant kinematic viscosity  and tribution of magnetic energy over di erent spherical harmonic the magnetic di usivity (r) which varies with the radial dis- degrees l at depth r. The Lowes spectrum R (r) in (4) is a special tance r. The dynamical variables of velocity, magnetic field, case of F (r) under the condition j = 0 (which is only true near entropy, density and pressure (u; B; S; ; p) are governed by the the planetary surface). The change in behaviour of F (r) along non-dimensional equations, r indicates varying dynamics in di erent regions. Comparing F (r) to R (r) gives further insights into the di erent physics in Ek Du EkRaPm dT l l 0 0 + 2zˆ  u = r S rˆ these regions. Pm Dt Pr dr (10a) In the next section, we describe our model for Jupiter’s dy- 1 F + (r B) B + Ek ; namo. In section 3, we first introduce the magnetic energy spec- ¯  ¯ trum and discuss how its behaviour changes with depth. We @B then show that a dynamo radius can be identified from a transi- = r (u B)r (r B); (10b) @t tion in the spectral slope. In section 4, we look at the relation- DS Pm ship between the dynamo radius and the Lowes radius. We then ¯ ¯ ¯T = r ( ¯TrS ) Dt Pr finish with a discussion on the di erences between results from (10c) Pm Pr 1 our model and observation. +  ¯T H + Q + Q ; S  J Pr RaPm Ek r ( ¯u) = 0; (10d) 2. A model of Jupiter’s dynamo r B = 0; (10e) Numerical models of Jupiter’s dynamo have recently been developed to study the magnetic field and internal flow of the gi- together with the equation of state ant planet (Jones et al., 2011; Jones, 2014; Gastine et al., 2014). The model used in the present study is developed and described @ @  ¯ dT 1 d ¯ d = dS + d p = dS d p: (11) in detail by Jones (2014), though here the range of parameters @S @ p g ¯ dr  ¯g ¯ dr p S has been extended to get further into the strong-field dynamo In deriving the model, the Lantz and Fan (1999) formulation regime (Dormy, 2016). We briefly summarise it here. of the anelastic approximation (Braginsky and Roberts, 2007) has been employed about a spherically symmetric, hydrostatic 3 ¯ and adiabatic basic state (; ¯ p ¯; S ). This simplifies the system to involve the dynamics of a single thermodynamical variable S = S S . In (10a), zˆ and rˆ are unit vectors along the rotation 0 0 0 axis and the radial direction respectively,  = p = ¯ +  is a generalised pressure combining disturbance pressure p = p p ¯ and disturbance gravitational potential  . In (11), g ¯ is the gravitational acceleration due to  ¯. The system is forced by a constant entropy source H modelling the secular cooling of the planet. The dissipative terms are: @u @ @u 2 @ j i F =  ¯ + ( ¯r u); (12a) ;i @x @x @x 3 @x j i j i j=1 X 2 1 @u @u 2 Q =  ¯ + (r u) ; (12b) 2 @x @x 3 i j i; j=1 Q = jr Bj : (12c) Let T ,  and  be the dimensional values of T ,  ¯ and , re- spectively, at the midpoint of the shell and S be the entropy drop across the thickness L = r r of the shell. Our equa- out in tions are non-dimensionalised using the unit of length L, time L = and magnetic field , where  is the perme- 0  0 ability of free space. The dimensionless numbers in (10) are defined as: Figure 3: Simulation at Pm = 10. (a) Dipole-dominated radial magnetic field T L S at r = r . (b) Radial magnetic field at r = 0:8r has larger magnitude and out J Ra = ; Ek = ; Pr = ; Pm = : (13) more small-scale structures. (c) Left: zonal velocity u at r = r showing the S S   out prograde equatorial jet. Right: zonal and time averaged u . The boundary condition for u is no-slip at r and stress-free at in r . At both the inner and outer boundaries, it is electrically out insulating and S is fixed at a constant value. The initial con- 2.3. Simulation parameters and electric current profile ditions are u = 0 with a small perturbation in B and S . The For the rest of this paper, the following parameters are kept spectrum of the initial magnetic perturbation is narrow-banded 7 5 fixed: r =r = 0:0963, Ra = 2  10 , Ek = 1:5  10 and in out with 8  l  10 and thus has no dipole component. Pr = 0:1. Jupiter is believed to have a strong-field dynamo, i.e., its magnetic field strongly influences the flow. Thus, we are in- 2.2. Hyperbolic electrical conductivity profile terested in cases of large magnetic Prandtl number Pm which In applying the anelastic convective system described above produce a strong-field dynamo (Dormy, 2016). Specifically, to model the Jovian dynamo, we need to provide an equilib- we investigate the e ects of Pm by comparing simulations with rium state and a conductivity profile that represent the thermo- Pm = 10 and Pm = 3. The case of Pm = 3 has previously been dynamic and transport properties inside Jupiter. French et al. studied in detail by Jones (2014). Results here are presented in (2012) have calculated the material properties of a hydrogen- dimensionless units unless otherwise stated. helium-water mixture under Jupiter-like condition using den- Figure 3 shows snapshots of magnetic and velocity fields sity functional theory. Here, we use the same equilibrium den- from the Pm = 10 simulation. The equatorial zonal jet and the sity  ¯(r) and temperature T (r) profile as in Jones (2014) which dipolar nature of the radial magnetic field, both near the surface, are smooth interpolations to the J11-8a data in French et al. are obvious. With the conductivity (r) increasing sharply with (2012). For the magnetic di usivity (r), we consider the fol- depth, we are interested in the radial dependence of the dynamo lowing hyperbolic model: action. We first examine the average electric current j at a rms (ln  + c r + c )(ln  + c r + c ) = c (14) 1 2 3 4 5 depth r, defined as, with five parameters. The values used for these parameters are D E 2 2 j (r) = j j(r; ; ; t)j sin dd: (15) the same as in Jones (2014) which give a good fit to the J11-8a rms data. Figure 2 shows the corresponding electrical conductivity (r) = 1= . In the interior, (r) decreases roughly linearly Above, j = r  B, hi indicates time average over a statisti- and reaches about one-fifth of its maximum value at r = 0:8r . cal steady state and the integral is over a spherical surface of This is unlike some of the previous studies (Gastine et al., 2014; radius r. Figure 4 shows that j (r) behaves very similarly for rms Glatzmaier, 2018) in which the electrical conductivity is taken both Pm = 3 and Pm = 10. In the interior where strong mag- to be constant below a certain depth. Dietrich and Jones (2018) netic field is being generated, j (r) increases slightly with r rms studied a wide range of profiles for (r) by varying c in (14) and peaks at around r = 0:7r even though (r) is monotoni- 2 J and found a diversity of magnetic field morphologies. cally decreasing in r. Approaching the surface, j (r) follows rms 4 Figure 5: Time evolution of selected modes of the magnetic energy spectrum F (r ; t) at the outer boundary r = r for the case of Pm = 10. The dipolar l out out l = 1 mode (dashed line) eventually becomes dominant as the system reaches a statistical steady state at about t = 0:05. Table 1 For simulations at two di erent Pm and Juno observation (with the two fitting ranges in Fig. 1(a)): (r ) is the spectral slope of F (= R at r ) measured at Figure 4: Average current j , defined in (15), at di erent depth r for Pm = 10 out l l out rms the outer boundary r and is measured inside the upper dynamo region and Pm = 3. The electrical conductivity (r) from Fig. 2 is also plotted on the out dyn as discussed below (21). r is the Lowes radius in (7) and r is the dy- right axis. lowes dyn namo radius in (22). Rm(r) is the depth-dependent magnetic Reynolds number defined in (25) and r is where Rm = 1. r = 6:9894 10 m. Rm J the trend of (r) and drops quickly and smoothly to negligi- (r ) r =r r =r Rm(r ) r =r out dyn lowes J dyn J dyn Rm J ble value, indicating the cessation of dynamo action. While the Pm = 10 0.072 0.024 0.883 0.907 156 0.939 variation of j (r) with r certainly signifies di erent dynamics Pm = 3 0.089 0.035 0.865 0.900 111 0.934 rms Juno (l  2) 0.109 ? 0.845 ? ? ? at di erent depth, Fig. 4 does not locate a characteristic depth Juno (l  5) 0.162 ? 0.796 ? ? ? that represents the top of the dynamo region. It is not obvious from the profile of j (r) at what depth the electric current be- rms comes large enough to generate a significant magnetic field. In interpreted as the average magnetic energy per spherical har- the next section, we show that a dynamo radius can be identified monic degree l (Maus, 2008). Note that F is calculated from using the magnetic energy spectrum. the full field B. On the other hand, R is calculated from the scalar potential in (2). Nevertheless, in a current-free region, 3. Magnetic energy spectrum the two are identical. Hence, We again consider the average magnetic energy on a spher- F (r) = R (r) if j (r) = 0: (19) l l rms ical surface given by (1). Unlike in the derivation of the Lowes spectrum R where the magnetic field is assumed to be potential, l 3.1. F (r) at the planetary surface here we make no such assumption and expand B in terms of a We first examine F (r) at the outer boundary r = r of l out m m m set of vector spherical harmonics fY ; ; g (see Appendix l l l the spherical shell. Figure 5 shows the time evolution of a few A), modes of F (r ; t) from the Pm = 10 simulation. The devel- l out 1 l X X m m m opment of the dipolar l = 1 mode can clearly be seen as it B = q Y + s + t  : (16) lm lm lm l l l outgrows all other modes. l=1 m=l Since the electric current is negligible at r , F (r ) = out l out The expansion coecients are generally function of r and t. R (r ) in our simulations, see for example Fig. 6 for the case l out Substituting (16) into (1), we get of Pm = 10. Figure 1(c) plots F from the Pm = 10 and Pm = 3 simulations as well as R calculated from the Juno data JRM09. All three spectra have been continued to r = r using (5) and 2 E (r; t) = F (r; t) (17) 0 B l J l=1 normalised. The key qualitative di erence between the spec- tra from our two simulations is that F (r ) for Pm = 3 displays l J where the magnetic energy spectrum is a clear exponential decay for all l (excluding l = 1) while for Pm = 10, the spectrum is roughly flat at small l and only starts 4 3 m;0 2 2 2 to decay exponentially for l & 5. In this respect, the Pm = 10 F (r; t) = jq j +js j +jt j : (18) l lm lm lm (2l + 1) spectrum is similar to the Juno spectrum. However, the decay m=0 rate of F for Pm = 3 is faster and slightly closer to the Juno We are mainly interested in the time-averaged spectrum F (r) = observed value. Fitting the range 5  l  40 yields the values hF (r; t)i , with the time-averaging done after the system has l t of r shown in Table 1 for Pm = 10 and Pm = 3. lowes reached a statistical stationary state. Roughly, F (r) can be 5 Figure 7: Spectral slope (r) of the magnetic energy spectrum F (r) and the spectral slope (r) of the Lowes spectrum R (r) in the Pm = 10 simulation. At the dynamo radius r , diverges from and remains more or less constant dyn about inside a large part of the dynamo region. The Lowes radius r is dyn lowes Figure 6: Time-averaged magnetic energy spectrum F (r) (solid lines), defined where the downward continued R is flat, (r ) = 0. l lowes in and below (17), and Lowes spectrum R (r) (circles) at di erent depth r for the Pm = 10 simulation. R (r) are obtained from the time-averaged R (r ) l l out using (5). from R (r). The main observation in Fig. 6 is that the shape of F (r) becomes independent of r, as discussed in the previous section, at essentially the same depth r . This implies that dyn 3.2. F (r) at di erent depth r r is the boundary below which electric current becomes im- dyn We now look at the magnetic energy spectrum in the inte- portant and the dynamics of the system is altered. We therefore rior of the spherical shell. We focus on the case of Pm = 10. identify r as the top of the dynamo region, or the dynamo dyn The solid lines in Fig. 6 show F (r) at di erent depth r. The key radius. feature is the transition of F (r) through three di erent stages Figure 7 vividly illustrates the discussion in the previous as r varies. We have seen in the previous section the steep ex- paragraph by plotting (r) together with the Lowes spectral ponential decay of F with l (for l & 5) at the surface. As we slope (r), defined analogously to (6), as a function of r. From move below the surface, Fig. 6 shows F (r) maintains such ex- (5), we have ponential decay, out (r) = (r ) 2 log ; (21) out log F (r)  (r)l; (20) where (r ) = (r ). Note how (r) diverges from (r) and out out levels o to the value at r = r . The sharpness of the dyn dyn but the spectral slope (r) decreases rapidly with r inside the transition allows for a meaningful definition of r . Figure 7 layer of 0:91r . r  r and F (r) has become rather shal- dyn J out l also provides a quantitative way to determine r . Fitting a dyn low at 0:91r . As we delve further into the interior, quite re- horizontal line through 0:7r < r < 0:86r yields the value of J J markably, F (r) remains shallow as its shape, and hence (r), = 0:024. We then obtain the dynamo radius r from (21) dyn dyn becomes more or less invariant over the substantial region of using the relation 0:55r . r . 0:91r . This clearly indicates a shift in the dy- J J (r ) = : (22) dyn dyn namics near 0:91r . Finally, in the deep interior and close to the core, F (r) decays super-exponentially and the magnetic field This gives r = 0:907r for the Pm = 10 simulation. The dyn J is dominated by large scales. Boussinesq geodynamo models values of the various spectral slopes and characteristic radii for which compute the magnetic energy spectrum inside the core Pm = 10 are summarised in Table 1. also show a spectrum that decays exponentially with l for l  5, Spacecraft missions can only measure the spectral slope at e.g. Christensen et al. (1999). the planetary surface, from which r is calculated using the lowes white source assumption discussed in section 1. The question 3.3. A dynamo radius is then how well can r predict the actual dynamo radius lowes The significance of the change in behaviour of F (r) be- r . In our simulations, r is where the dashed line in Fig. 7 dyn lowes comes clear when we compare F (r) to the Lowes spectrum intersects the horizontal axis, i.e. (r ) = 0 in (21). This is lowes R (r) at the same depth. Recall that R (r ) = F (r ) at the because by definition, the downward continued R (r) becomes l l out l out surface. We now downward continue R (r ) using (5) (with flat at r = r , at least for the range of l where R (r ) is fitted l out lowes l out a = r ) to obtain R (r) at di erent r, which we plot as circles to obtain (r ). It is clear from Fig. 7 that generally out l out in Fig. 6. Near the surface, F (r) and R (r) are essentially in- l l r  r : (23) distinguishable because electric current is negligible there. As lowes dyn r decreases and reaches some depth r , F (r) starts to deviate dyn l 6 3.4. E ects of Pm We now compare the simulation at Pm = 10 to the one at Pm = 3. We have already discussed the di erences in the magnetic energy spectrum at the surface in section 3.1. The slightly steeper F (r ) for Pm = 3 means the magnetic field l out has less small scales. Generally, results from the two simula- tions are qualitatively similar. The Pm = 10 simulation has stronger magnetic fields, so the Elsasser number, which is pro- portional to the root-mean-squared magnetic field over the do- main, is about an order of magnitude larger than that of the Pm = 3 simulation, closer to the high value expected in Jupiter. In the Pm = 3 simulation, it was noted in Jones (2014) that the Lorentz force in the interior mostly suppresses the inter- nal di erential rotation in the metallic hydrogen region. At Pm = 10 the stronger magnetic field means the internal di er- ential rotation is even more strongly suppressed, see Fig. 3(c), so that strong zonal flow occurs only in the molecular region near the equator. However, although the zonal flow is confined to the molecular region, its magnitude is about double that of the Pm = 3 simulation. The convective velocity in the Pm = 10 simulation is also about double that in the Pm = 3 simulation, so the magnetic Reynolds number is about twice that of the Pm = 3 simulation. While it is not computationally possible to achieve the parameters believed to operate in Jupiter, the in- crease in Pm has moved the simulation results in the direction of more realistic values. Figure 4 shows the electric current is roughly three times smaller in the Pm = 3 case. The spectral slopes for Pm = 3 Figure 8: Selected modes of the magnetic energy spectrum F (r) and the Lowes display the same trend as in Fig. 7. The values of (r ) and out spectrum R (r) as a function of depth r in the simulation at Pm = 10. (a) F (r) l l over the whole spherical shell. (b) F (r) (solid lines) and R (r) (dotted lines) together with that of r and r are given in Table 1. l l dyn lowes dyn near a region about the dynamo radius r . dyn A steeper spectrum at the surface means a bigger (r ). At out the same time, we see that also increases. The net result dyn is that the dynamo radius r is only marginally less than that dyn Comparing the values shown in Table 1, we see that for Pm = of Pm = 10. On the other hand, r drops more significantly lowes 10, r is about 3% less than r . The di erence stems from lowes dyn which makes r less accurate as a predictor of r . These lowes dyn being fairly small but not exactly zero. Comparing F to dyn l results suggest that for a dynamo with a larger Pm, the magnetic R near r in Fig. 6 again shows the white source assump- l lowes energy spectrum F (r) in the upper part of the dynamo region tion is only approximate and F (r) never becomes exactly flat. is closer to being ‘white’. As a consequence, the Lowes radius Nonetheless, helped by the steep decrease of (r) with r shown gives a better prediction to the dynamo radius. in Fig. 7, we still have a close agreement between r and lowes r . Figure 3(b) shows a snapshot of the radial magnetic field dyn at r = 0:8r where (0:8r )  . 4. Discussion and conclusion J J dyn Figure 8 reveals further details about the radial dependence The electrical conductivity in Jupiter varies from being neg- of the magnetic energy spectrum by plotting a selected number ligible at the surface to a very high value in the interior. It of modes of F (r) as a function of r. The first few modes (l . 5) thus raises the question about the depth at which dynamo action are exceptional as they varies irregularly with r while the rest starts. In this paper, we consider the magnetic energy spectrum of the modes are well represented by l = 20 in Fig. 8(a). The F (r) at depth r in a numerical model of Jupiter’s dynamo. For l = 1 mode has the largest magnitude at all r except for a region l & 5, the magnetic energy spectrum decays exponentially with around 0:8r where it is overtaken by l = 2, see also Fig. 3(b). l, log F (r)  (r)l. We find that a sharp transition in (r) Interestingly, this is also the region in which j (r) peaks and 10 rms can be used to identify a dynamo radius r and this dynamo dyn (r) remains virtually constant at . We also see that all dyn radius can be reasonably predicted by the Lowes radius r lowes non-dipolar modes are strongly damped just above the dynamo as discussed in section 3.3. The situation is in fact rather sim- radius r leading to the dipole-dominated field observed at dyn ple as illustrated in Fig. 7. The two characteristic radii r and dyn the surface. Figure 8(b) zooms into a region about r = r and dyn r are controlled by two spectral slopes: (r ) which is ob- lowes out shows, for the large-scale modes, how F (r) deviates from R (r) l l servable at the surface and which measures the deviation dyn as one moves from the current-free outer layer into the dynamo from the white source hypothesis near the top of the dynamo region. 7 However, Rm does give a somewhat smaller r  0:930r r Rm J and Rm (r )  O(10) for Pm = 10. The three magnetic r dyn Reynolds numbers in (25) all use the shell thickness L as the typical length scale. An alternative is to use the magnetic dif- fusivity scale height d (r) = (d=dr) . Dietrich and Jones (2018) and Wicht et al. (2019) have examined, among other things, the characterisation of Jovian dynamo models using dif- ferent magnetic Reynolds numbers. Since d  L, we expect Rm  Rm  O(10) and Rm  O(1) at r in our simulations r dyn if the definition of Rm based on d (r) is used. On the other hand, numerical dynamo models cannot achieve the large value of Rm expected in the interior of Jupiter. If we adopt the velocity of 2 1 U = 10 m s which has been estimated for Jupiter’s interior, see e.g. Jones (2014), and use our values of L and , then the value of r at which Rm = U L= = 1 increases to r = 0:957r . We have estimated the Lowes radius from the Juno data in Fig. 1. We find that it has fairly large uncertainties depending on the range of l used in the linear fitting. However, if the zonal components are omitted we find a closer linear fit for l  6 as did Langlais et al. (2014) for the geomagnetic data. The situ- ation in Jupiter could be similar to our simulation at Pm = 10 Figure 9: Depth-dependent magnetic Reynolds numbers, defined in (25), based where a clean exponential decay in the spectrum emerges only on di erent velocity scales for Pm = 10 and Pm = 3. The two vertical dot- dashed lines indicate the locations of r and r given in Table 1. dyn lowes at larger l. This is in contrast to the case of Pm = 3, which has weaker flow and magnetic field. We anticipate further data col- lected by the ongoing and future flybys will extend the range of region. Varying moves the horizontal dot-dashed line in dyn the Lowes spectrum in Fig. 1 and hence provide a more reliable Fig. 7 up and down while changing (r ) shifts the dashed out estimate of r . lowes curve left and right. These determine the location of r and dyn Despite uncertainties in the data, Table 1 shows that r lowes r as well as the relative distance between them. Notice that lowes from the Juno observation is clearly smaller than in both of our the dashed curve, given by (21), is essentially a straight line for simulations. Nevertheless, its implication on the location of the r r  r , out out dynamo radius is not clear. Whether r gives a good esti- lowes 2r mate on r relies on the white source hypothesis, which may dyn (r)  2 + (r ): (24) out or may not be valid in Jupiter. However, the Pm = 10 simula- out tion, which we believe is closer to Jupiter conditions than the We find that in our two simulations at Pm = 10 and Pm = 3, Pm = 3 simulation, has a smaller than the Pm = 3 sim- dyn (r ) and change in such a way that leaves the dynamo out dyn ulation, suggesting that Jupiter’s magnetic field might be close radius fairly insensitive to Pm. Incidentally, at r = r , the dyn to white near the top of the dynamo region. In our simulations, electrical conductivity (r) has dropped by two orders of mag- a steeper spectrum observed at the surface tends to be accom- nitude from its maximum at the inner boundary. panied by steeper spectra in the interior and consequently r dyn Figure 9 plots as a function of depth three di erent magnetic could be shallower than r by a fair amount. The present lowes Reynolds numbers, each based on a di erent velocity scale, results suggest that the true dynamo radius likely lies above the Lu (r) Lu (r) Lu (r) Lowes radius. ;rms rms r;rms Rm = ; Rm = ; Rm = ; Irrespective of its relation to the dynamo radius, the Lowes (r) (r) (r) (25) radius is a property of the magnetic field at the surface of for our two Jovian dynamo simulations, reverting to dimen- Jupiter. The smaller r of the Juno data stems from a steeper lowes sional units. Here, u , u and u are the root-mean- spectrum at the planetary surface, implying Jupiter’s magnetic rms r;rms ;rms squared values of the total, radial and zonal velocity, respec- field has less small scales than that in our model. This is slightly tively, over a spherical surface of radius r. All three magnetic surprising as the flow is believed to be more vigorous in Jupiter Reynolds numbers vary weakly in the interior and then decrease than in our simulations, because the simulations have enhanced sharply to a negligible value near the surface. We also see that di usion coecients to maintain numerical stability. We should Rm  Rm for r & r indicating the zonal flow becomes point out again that r for the simulations in Table 1 are dyn lowes dominant. This is consistent with the depth of the equatorial derived from time-averaged spectra while the Juno observa- zonal jet estimated roughly from figures such as Fig. 3. The tion essentially provides only a snapshot of Jupiter’s magnetic values of Rm at r and the depth r at which Rm = 1 are field. Figure 10 plots the instantaneous Lowes radius r (t) dyn Rm lowes given in Table 1. These values suggest that r determined obtained from the time-dependent spectrum F (r ; t) in the Rm l out by the criterion Rm = 1 is generally larger than r estimated Pm = 10 simulation. The case of Pm = 3 shows similar spread dyn from F (r). Using Rm instead of Rm gives similar estimates. about the mean value. We argue that the di erence between 8 Acknowledgements The authors are supported by the Science and Tech- nology Facilities Council (STFC), ‘A Consolidated Grant in Astrophysical Fluids’ (grant numbers ST/K000853/1 and ST/S00047X/1). This work used the DiRAC@Durham facility managed by the Institute for Computational Cosmology on be- half of the STFC DiRAC HPC Facility (www.dirac.ac.uk). The equipment was funded by BEIS capital funding via STFC cap- ital grants ST/P002293/1, ST/R002371/1 and ST/S002502/1, Durham University and STFC operations grant ST/R000832/1. DiRAC is part of the National e-Infrastructure. Figure 10: Instantaneous Lowes radius r (t) obtained from the time- lowes dependent magnetic energy spectrum F (r ; t) at the outer boundary for Pm = l out 10. The dashed line is the time-averaged value over the statistical steady state Appendix A. Vector spherical harmonics given in Table 1. Following Barrera et al. (1985) but using the Schmidt semi- normalised associated Legendre polynomials P in the defini- simulations and observation is significant even when statistical tion of the spherical harmonics, fluctuation is taken into account. jmj The di erence in r between our numerical model and m im lowes Y (; ) = P (cos )e ; (A.1) l l observation raises several questions and suggests possible av- enues for future research. The results presented here are spe- we define three vector spherical harmonics: cific to the electrical conductivity profile (r) in (14) which m m is based on data from theoretical ab initio calculation. Could Y (; ) = Y rˆ; (A.2a) l l the deeper Lowes radius in observation mean the actual electri- m m (; ) = p rrY ; (A.2b) cal conductivity inside Jupiter is smaller than predicted? It is l l l(l + 1) worth studying how the magnetic energy spectrum responds to m m (; ) = rˆ  ; (A.2c) l l perturbations in (r). On a related note, the magnetic Reynolds number Rm in our simulations is of the order 10 at its max- which form an orthogonal basis for all square-integrable vector imum, much lower than an estimated value of 10 in Jupiter fields on the unit sphere. The (semi-)normalisation condition is (Jones, 2014). The puzzle here is that increasing Rm in the model will likely increase r rather than reduce it and thus lowes 0 4 m m 0 0 Y  (Y 0 ) sin dd = (2  )  ; (A.3) m;0 ll mm l l move it further away from the Juno value. While current com- 2l + 1 puting resources prevent us from reaching a much larger Rm, m m with similar expressions for and  . investigating the trend of the dynamics in the neighbourhood l l of a smaller attainable Rm, possibly by changing (r), could provide valuable insights. References In our simulations, the system is forced by the constant en- Backus, G., Parker, R., Constable, C., 1996. Foundations of Geomagnetism. tropy source H in (10c) and we employ a constant entropy Cambridge University Press, Cambridge. boundary condition. Could our numerical setup tend to pro- Barrera, R. G., Estev ´ ez, G. A., Giraldo, J., 1985. Vector spherical harmonics duce extra small scales that lead to the shallower spectra shown and their application to magnetostatics. Eur. J. Phys. 6, 287–294. in Fig. 1(c)? In geodynamo simulations, boundary conditions Bolton, S., Levin, S., Bagenal, F., 2017. Juno’s first glimpse of Jupiter’s com- plexity. Geophys. Res. Lett. 44, 7663–7667. can significantly a ect the dynamics (Sakuraba and Roberts, Braginsky, S. I., Roberts, P. H., 2007. Anelastic and Boussinesq approxima- 2009; Dharmaraj and Stanley, 2012). It is important to assess tions. In: Gubbins, D., Herrero-Bervera, E. (Eds.), Encyclopedia of Geo- the robustness of the present results and examine their depen- magnetism and Paleomagnetism. Springer Netherlands, pp. 11–19. dence on boundary conditions and the form of forcing. Cao, H., Stevenson, D. J., 2017. Zonal flow magnetic field interaction in the semi-conducting region of giant planets. Icarus 296, 59–72. The formation of a stably stratified layer just under the Christensen, U., Olson, P., Glatzmaier, G. A., 1999. Numerical modelling of the molecular layer due to ‘helium rain’ (Stevenson and Salpeter, geodynamo: a systematic parameter study. Geophys. J. Int. 138, 393–409. 1977) has been proposed to explain the near-axisymmetric Connerney, J. E. P., Acuan, ˜ M. H., Ness, N. F., Satoh, T., 1998. New models of magnetic field of Saturn (Stevenson, 1982; Dougherty et al., Jupiter’s magnetic field constrained by the Io flux tube footprint. J. Geophys. Res. 103, 11929–11939. 2018). Although helium rain is more probable to occur in Sat- Connerney, J. E. P., Kotsiaros, S., Oliversen, R. J., Espley, J. R., Joergensen, urn, it cannot be ruled out for Jupiter (Wahl et al., 2017; Debras J. L., Joergensen, P. S., Merayo, J. M. G., Herceg, M., Bloxham, J., Moore, and Chabrier, 2019). It would be interesting to see the e ects K. M., Bolton, S. J., Levin, S. M., 2018. A new model of Jupiter’s magnetic field from Juno’s first nine orbits. Geophys. Res. Lett. 45, 2590–2596. of such a stable layer on the Lowes radius as it displaces the dy- Debras, F., Chabrier, G., 2019. New models of Jupiter in the context of Juno namo action deeper into the interior. This is perhaps the most and Galileo. Astrophys. J. 100, 872. natural way to explain the surprisingly low value of r in the lowes Dharmaraj, G., Stanley, S., 2012. E ect of inner core conductivity on planetary Juno data. dynamo models. Physics of the Earth and Planetary Interiors 212-213, 1–9. 9 Dietrich, W., Jones, C. A., 2018. Anelastic spherical dynamos with radially variable electrical conductivity. Icarus 305, 15–32. Dormy, E., 2016. Strong-field spherical dynamos. J. Fluid Mech. 789, 500–513. Dougherty, M. K., Cao, H., Khurana, K. K., Hunt, G. J., Provan, G., Kellock, S., Burton, M. E., Burk, T. A., Bunce, E. J., Cowley, S. W. H., Kivelson, M. G., Russell, C. T., Southwood, D. J., 2018. Saturn’s magnetic field revealed by the Cassini Grand Finale. Science 362 (6410), eaat5434. French, M., Becker, A., Lorenzen, W., Nettelmann, N., Bethkenhagen, M., Wicht, J., Redmer, R., 2012. Ab initio simulations for material properties along the Jupiter adiabat. Astrophys. J. Suppl. 202, 5 (11pp). Gastine, T., Wicht, J., Durate, L. D. V., Heimpel, M., Becker, A., 2014. Ex- plaining Jupiter’s magnetic field and equatorial jet dynamics. Geophys. Res. Lett. 41, 5410–5419. Glatzmaier, G. A., 2018. Computer simulations of Jupiter’s deep internal dy- namics help interpret what Juno sees. PNAS 115, 6896–6904. Helled, R., 2018. The interiors of Jupiter and Saturn. Oxford Research Ency- clopedias: Planetary Science. Hori, K., Teed, R. J., Jones, C. A., 2019. Anelastic torsional oscillations in Jupiter’s metallic hydrogen region. Earth Planet. Sci. Lett., accepted. Jones, C. A., 2014. A dynamo model of Jupiter’s magnetic field. Icarus 241, 148–159. Jones, C. A., 2018. Jupiter’s magnetic field revealed. Nature 561, 36–37. Jones, C. A., Boronski, P., Brun, A. S., Glatzmaier, G. A., Gastine, T., Miesch, M. S., Wicht, J., 2011. Anelastic convection-driven dynamo benchmarks. Icarus 216, 120–135. Langlais, B., Amit, H., Larnier, H., Thebault, ´ E., Mocquet, A., 2014. A new model for the (geo)magnetic power spectrum, with application to planetary dynamo radii. Earth Planet. Sci. Lett. 401, 347–358. Lantz, S. R., Fan, Y., 1999. Anelastic magnetohydrodynamic equations for modeling solar and stellar convection zones. Astrophys. J. Suppl. 121, 247– Liu, J., Goldreich, P. M., Stevenson, D. J., 2008. Constraints on the deep-seated zonal winds inside Jupiter and Saturn. Icarus 196, 653–664. Lowes, F. J., 1974. Spatial power spectrum of the main geomagnetic field and extrapolation to the core. Geophys. J. R. Astron. Soc. 36, 717–730. Mauersberger, P., 1956. Das mittel der energiedichte des geomagnetischen ¨ ¨ ¨ hauptfeldes an der erdoberflache und seine sakulare anderung. Gerlands Beitr. Geophys 65, 207–215. Maus, S., 2008. The geomagnetic power spectrum. Geophys J. Int. 174, 135– Moore, K. M., Yadav, R. K., Kulowski, L., Cao, H., Bloxham, J., Connerney, J. E. P., Kotsiaros, S., Jørgensen, J. L., Merayo, J. M. G., Stevenson, D. J., Bolton, S. J., Levin, S. M., 2018. A complex dynamo inferred from the hemispheric dichotomy of Jupiter’s magnetic field. Nature 561, 76–78. Ridley, V. A., Holme, R., 2016. Modeling the Jovian magnetic field and its secular variation using all available magnetic field observations. J. Geophys. Res. Planets 121, 309–337. Sakuraba, A., Roberts, P. H., 2009. Generation of a strong magnetic field using uniform heat flux at the surface of the core. Nature Geoscience 2, 802–805. Stevenson, D. J., 1982. Reducing the non-axisymmetry of a planetary dynamo and an application to Saturn. Geophys. Astrophys. Fluid Dyn. 21, 113–127. Stevenson, D. J., Salpeter, E. E., 1977. The dynamics and helium distribution in hydrogen-helium fluid planets. Astrophys. J. Suppl. 35, 239–261. Wahl, S. M., Hubbard, W. B., Militzer, B., Guillot, T., Miguel, Y., Movshovitz, N., Kaspi, Y., Helled, R., Reese, D., Galanti, E., Levin, S., Connerney, J. E., Bolton, S. J., 2017. Comparing Jupiter interior structure models to Juno gravity measurements and the role of a dilute core. Geophys. Res. Lett. 44, 4649–4659. Weir, S. Y., Mitchell, A. C., Nellis, W., 1996. Metallization of fluid molecular hydrogen at 140GPa (1.4Mbar). Phys. Rev. Lett. 76, 1860–1863. Wicht, J., French, M., Stellmach, S., Nettelmann, N., Gastine, T., Durate, L., Redmer, R., 2018. Modeling the interior dynamics of gas planets. In: Luhr ¨ , H., Wicht, J., Gilder, S. A., Holschneider, M. (Eds.), Magnetic Fields in the Solar System. Vol. 448 of Astrophysics and Space Science Library. Springer Cham, Ch. 2, pp. 7–81. Wicht, J., Gastine, T., Durate, D. V., 2019. Dynamo action in the steeply de- caying conductivity region of Jupiter-like dynamo models. J. Geophys. Res. Planets 124, 837–863. Wigner, E., Huntington, H. B., 1935. On the possibility of a metallic modifica- tion of hydrogen. J. Chem. Phys 3, 764–770.

Journal

Astrophysics – arXiv (Cornell University)

Published: May 18, 2019

There are no references for this article.