Deuterium Abundance from HD and CH 3 D Reservoirs in the Atmosphere of Jupiter

Transcription

1 1 Deuterium Abundance from HD and CH 3 D Reservoirs in the Atmosphere of Jupiter C. D. Parkinson Institut d Astrophysique de Paris-CNRS, Paris, France. L. Ben Jaffel Institut d Astrophysique de Paris-CNRS, Paris, France J. C. McConnell Department of Earth and Atmospheric Science, York University, Toronto, Canada Short title: DEUTERIUM ABUNDANCE IN THE JOVIAN ATMOSPHERE

2 2 Abstract. CH 3 D is an isotopic tracer in the deep Jovian atmosphere and susceptible to transport and chemical effects. It is expected that the tropospheric ([D]/[H]) CH4 ratio calculated from data collected from various observations should be relatively invariable, yet previous determinations of this quantity in Jupiter have given results inconsistent with experimental error bars. This suggests that there may be a problem with the interpretion of some of the observations, or that the apparent CH 3 D column abundance is variable. We report on the effects of varying important parameters over this pressure regime on a standard reference atmosphere and discuss how this impacts the CH 3 D mixing ratio, CH 3 D fractionation, the ([D]/[H]) CH4 and D/H (= ([D]/[H]) H2 ) ratios and compare with the various CH 3 D and HD observations. We proceed by assuming that the in situ measurement of the D/H ratio obtained from the Galileo Probe Mass Spectometer (GPMS) is correct which is then used as a boundary condition in the deep atmosphere. We then calculate the HD mixing ratio profile and a range of CH 3 D mixing ratio profiles using a 1-D eddy diffusion model with chemical rates given by Lécluse et al. (1996) which yields the isotropic enrichment factor, f(z), for each eddy diffusion profile, K(z). Our D/H value derived via the CH 3 D modelling from the deep atmosphere to the millibar pressure region reflects the GPMS D/H value initially assumed and makes our model self-consistent in the following sense: both the CH 3 D and HD reservoirs lead to the same D/H ratio when the CH 3 D chemistry and the atmospheric structure are

3 3 described in the way presented in this study. Moreover, using this technique allows for the first time a way to explain the discrepancies in the ([D]/[H]) CH4 ratio observations since CH 3 D was first detected on Jupiter nearly 30 years ago by offering a plausible link between the CH 3 D observations and upper tropospheric dynamical processes thus resolving this long standing issue.

4 4 Introduction In the early formation of the solar system, most of the deuterium in the protosolar nebula was in hydrogen as HD. Since the protosolar nebula was a dense medium, isotopic exchange of H and D could occur via reactions between HD and neutral species (Geiss and Reeves, 1981) forming secondary reservoirs of deuterium, viz., HDO, CH 3 D, NH 2 D, and HDS. Jupiter and Saturn are considered to be undisturbed deuterium reservoirs from the protosolar nebula (being free from production or loss processes; Hubbard and MacFarlane, 1980) and hence their atmospheres should reflect the abundance of H and D since the formation of the solar system 4.5 Gyrs ago (Owen et al., 1986). The D/H ratio is an important parameter for scenarios describing the evolution of the solar system. For a long time the only inferences of this ratio for the giant planets were made by studying the HD line in the visible range or CH 3 D transitions in the near or mid IR range. HD observations in the visible range are subject to observational problems and difficulties in modelling line formation and radiative transfer for such weak lines in a cloudy atmosphere (Fegley and Prinn, 1988). The first measurements of CH 3 D in Jupiter s atmosphere were at 4.5 to 4.8 µm in the IR spectral region which yielded D/H = 5.1(±2.2) 10 5 (Beer and Taylor, 1973). Some other CH 3 D spectroscopic observations are: Bjoraker et al. (1986) found D/H = 1.2(±0.5) 10 5 from 5 µm airborne observations; Gautier and Owen (1989) obtained D/H = 2.1(±0.6) 10 5 by analysing Voyager IR data at 7.7 µm combined with the airborne observations by Bjoraker et al. (1986) to get CH 3 D/H 2 ; and a reanalysis of the Voyager IR data by

5 5 Carlson et al. (1993) yielded D/H = 3.6(±0.5) These measurements are limited by uncertainties in the scattering processes that arise in the line formation and require the determination of the isotopic enrichment factor, f(z), which depends upon the CH 3 D/CH 4 ratio (e.g. Encrenaz et al., 1999). Hence, for a true comparison between these observations, these D/H ratios need to be normalised to a standard reference f(z) and CH 4 mixing ratio (cf. Table ). Recently, new spectroscopic measurements of the D/H ratio have been obtained by analysing the far infrared spectrum recorded by the Infrared Space Observatory (ISO) (Encrenaz et al., 1996; 1999). Using the 5 µm region of the short wavelength spectrometer (SWS) which is sensitive to the 2-5 bar pressure region and the ν 2 band of CH 3 D, they obtained a CH 3 D volume mixing ratio Together with an assumed fractionation factor of 1.37 this gave a D/H ratio 2.2(±0.5) A similar D/H ratio was also obtained using the HD R(2) and R(3) lines at about 37.7 µm and 28.5 µm, H 2 S(0) and S(1) quadropole lines at 17.1 and 28.2 µm, methane ν 4 band at 7.7 µm, and the CH 3 D ν 6 band at 8.6 µm similtaneously on the ISO Short Wavelength Spectrometer (SWS) (Lellouch et al., 2001). This yielded a D/H ratio of 2.4(±0.4) 10 5 and 2.2(±0.7) 10 5 from the HD & H 2 data and the CH 3 D & CH 4 data, respectively. Additionally, in-situ measurements of the D/H ratio have been made using observations by the Galileo probe mass spectrometer (GPMS). The most recent estimate has been the analysis by Mahaffy et al. (1998) who give a value of 2.6(±0.7) 10 5 appropriate for the bar pressure region. These additional observations have also been summarised in Table.

6 6 Fegley and Prinn (1988) and Lécluse et al. (1996) have previously modelled CH 3 D and deuterium enrichment in the Jovian atmosphere using 1-D chemical and diffusion models. Fegley and Prinn (1988) used rates for isotopic exchange that were modelled from data on radical species, and predicted the fractionation factors for the deuterium exchange between HD and CH 3 D as a function of K(z). More recently, Lécluse et al. (1996) used published laboratory measurements of the isotopic exchange rate constant between CD 4 and H 2 to calculate f(z) between CH 4 and H 2 and CH 3 D abundances and employed a different transport model than that used by Fegley and Prinn (1988). The fractionation factor so far derived from the two models differ mainly because of a difference in obtaining the isotopic exchange rate below 700K. Fegley and Prinn (1988) used a theoretical expression adjusted to fit laboratory measurements made at 700K whereas Lécluse et al. (1996) have inferred their fit using laboratory measurements of isotopic exchange between hydrogen and methane molecules below 700K. The two expressions of the rate of isotopic exchange agree at 700K. This paper addresses the problem of the abundance of CH 3 D from the deep troposphere up to the mesosphere of Jupiter employing the assumption that isotopic exchange of D between CH 4 and H 2 represents the only chemistry in the deep atmosphere (following Lécluse et al. (1996)) and that vibrationally excited H 2 and HD chemistry in the thermosphere will constrain the CH 3 D upper boundary condition (Parkinson et al., 2000). We cast our results in terms of a standard reference model atmosphere, calculated by constructing a standard reference eddy diffusion profile based on various observations and theoretical modelling by other researchers.

7 7 Model Description Chemistry In this study, we calculate the isotopic exchange between H 2 and CH 4 in Jupiter s atmosphere assuming the following reversible reaction CH 4 + HD CH 3 D + H 2, k f, k r. (1) We use the rates given by Lécluse et al. (1996) for reaction (1) with k r = 0.25 ( e ( /T ) ), and k f = 0.5 k r α; α = e a, a = ( 1 T ( 1 T ) 3 ( ) T ) ( ) T In thermal equilibrium, isotopic exchange is dependent only on the temperature and is given by the isotopic fractionation factor, α(t), which is defined as the ratio of the D/H ratio in the deuterated species with the D/H ratio of H 2, viz., α(t) 1 [CH 3 D] 4 [CH 4 ] 1 [HD] 2 [H 2 ] = k f k r. (2) 1 2 and 1 4 are symmetry numbers which are a chemical property and the bracketed quantities represent number densities. Since the isotopes may not be in thermal equilibrium, a similar relation to α(t), termed the isotopic enrichment factor, f(z), may be defined by f(z) ([D]/[H]) CH4 /([D]/[H]) H2 = 1 [CH 3 D] 4 [CH 4 ] 1 [HD] 2 [H 2 ]. (3) α(t) and f(z) represent the tendency for D to be more tightly bound than H and therefore D becomes more concentrated in CH 4 than in H 2 (e.g. Gautier and Owen, 1989). While α(t) is only temperature dependent, the definition of f(z) allows for the

8 8 kinetics of reaction (1) and the vertical transport within of the atmosphere. Thus, it is not expected that isotopic equilibrium is obtained since the isotopes may not be in thermal equilibrium. In the deep atmosphere where temperatures exceed 1200 K kinetic, chemical time constants, τ chem, are sufficiently short that isotopic equilibrium is appropriate and f(z) α(t ). Higher in the atmosphere temperatures are lower and α(t ) increases. However, τ chem also dramatically increases so that the eddy diffusion time constant, τ K, is much smaller in comparison, and hence vertical mixing plays a more important role. Thus we may have a variation in the ([D]/[H]) CH4 ratio of the atmosphere depending on the value of K(z) in this region. Above the height where τ chem τ K kinetics will be too slow to change f(z) which is maintained by upward mixing. When this condition occurs, the critical temperature or height is often called the quench temperature or height, respectively (cf. Lécluse et al., 1996). This point typically resides below 100 bar with a value of 800K (Fegley and Prinn, 1988). The species volume mixing ratio, f i, used in this paper is calculated by solving the continuity equation for each species, i, (e.g., Chamberlain and Hunten, 1987) where the vertical flux is given by φ K i = Km df i dz. (4) K = K(z) is the vertical eddy diffusion coefficient (cm 2 s 1 ) which parameterizes macroscopic motions, such as the large scale circulation and planetary gravity waves, and m is the background number density.

9 9 Above 1 bar, we assume the ideal gas law is adequate, while below 1 bar, the ideal gas law becomes increasingly less valid in the deep atmosphere and the equation of state is more accurately represented by the van der Waals equation, which is valid down to 10 3 bars (Lécluse et al., 1996; for our purposes, a maximum pressure of 10 3 bars is adequate since we only need to go far enough down in the atmosphere such that f(z)=α(t ) 1). Lécluse et al. (1996) have published such temperature, T, pressure, p, and density, ρ profiles in the atmospheres of the giant planets and we have used an updated version (private communication: D. Gautier, 1999) in our calculations. We also considered the importance of a small variation of the temperature on the CH 3 D abundance using the diffential form of the van der Waal s equation of state for the number density. The atmosphere used is assumed to contain a helium mole fraction of (von Zahn and Hunten, 1996), for the methane mole fraction (Niemann et al., 1996), and an HD mixing ratio of (Mahaffy et al., 1998) at the lower boundary. It is further assumed that He, CH 4, and HD have a zero flux at 0.1 µbar, which is the upper boundary condition for each of these species in our model atmosphere. For CH 3 D we have used a fixed mixing ratio boundary condition of at 0.1 µbar (Parkinson et al., 2000) at the top of the atmosphere and assume thermochemical steady state at the bottom of the atmosphere which is at the 1000 bar level. Massie and Hunten (1982) and Carlson et al. (1993) treat ortho-para H 2 conversion in extensive detail, discussing how catalytic reactions between the free-radical surface sites of aerosol particles may affect this conversion in the tropopause region. This is of

10 10 possible interest since para hydrogen distribution and cloud opacities can be used along with the spatial variations of condensible species as tracers of the local and large-scale dynamics in the Jovian troposphere. However, following the analysis of Fegley and Prinn (1988), there is no compelling reason to indicate that cloud and aerosol particles catalytically alter the D/H ratios in CH 3 D in the troposphere of Jupiter and so we do not consider this effect in our calculations. Temperature Profile The temperature profile, T, is shown in Figure 1. Down to 24 bars T is determined by the Galileo Probe mass spectrometer (GPMS), as reported by Seiff et al. (1997). Below this we adopt the updated tropospheric temperature profile of Lécluse et al. (1996) (private communication: D. Gautier, 1999). Their T profile goes up to 1 bar, but the overlapping regions between 1 and 24 bar are equivalent in each case, thus ensuring no discontinuities between the two profiles. Eddy Diffusion Coefficient Figure 2 shows the eddy diffusion profiles we considered for this study. The standard value for the eddy diffusion coefficient at the homopause, K h cm 2 s 1, adopted is based on a reanalysis of the Voyager He 584 Å airglow data (Vervack et ( ) nbg (z al., 1995). From the mesosphere down to the tropopause, we use K(z) = K 0 ) γ 0 n bg (z) where K 0 = K h and n bg (z 0 ) is the background number density at level z 0. Gladstone et al. (1996) consider values of γ from 0.35 and 0.55, and choose a standard value of 0.45

11 11 in order to better fit the variation of C 2 H 6 in the lower stratosphere. Edgington et al. (1998; 1999) use γ = 0.6 to fit NH 3, C 2 H 2, and PH 3 Faint Object Spectrograph (FOS) observations. Variations of K(z) above the tropopause within these parameters effected neglible changes in our results, and so we have adopted γ = 0.5 as the standard reference, as illustrated in Figure 2. In the region of the tropopause, viz., the lower stratosphere and upper troposphere between 200 to 500 mbars, K(z) has been constrained by analysis of the ammonia and phosphine observations to be about 10 3 to 10 4 cm 2 s 1 (Strobel, 1973, 1977; Edgington et al., 1998, 1999; Atreya et al., 1999) and hence our standard reference values reflect the data for this region. In the region critical for these calculations, viz., the troposphere, there is not much information that would allow us to tightly constrain K(z) effectively in this region. Landry et al. (1991) show that the stratosphere may be connected to the troposphere by way of three-segment piecewise continuous eddy diffusion profile, where K(z) in the top segment is relatively constant and each of the two lower segments may have the form ( ) nbg (z K(z) = K ref ) γ ref n bg (z) and (nbg (z ref ) is the background density at level z ref ). For our model, the top segment corresponds to the tropopause region as previously described, the middle segment is the upper troposphere from the tropopause down to 20 bar, and the last segment the lower troposphere below 20 bar down to 10 3 bar. The choice of 20 bar in the middle segment is discussed in the following section. Assuming that the mixing length is about equal to the scale height H, Lécluse et al. (1996) connect the eddy diffusion coefficient with the vertical velocity, v, of a parcel of Jovian air using the relation K(z) = vh and obtain a nominal eddy diffusion profile that

12 12 varies with depth to about K(z) 10 9 cm 2 s 1 at the bottom. Smith (1998) showed that an effective length scale should be used instead of using the pressure scale height when deriving K(z). Fegley and Prinn (1988) consider a range of constant vertical eddy diffusion coefficients from 10 7 to 10 9 cm 2 s 1 throughout the entire atmosphere, but use a baseline value of K(z) = cm 2 s 1 for most of their modelling efforts. Landry et al. (1991) attempt to reproduce observations of CO at 5 bar and and constrain C 2 H 6 abundances in the upper troposphere by considering a variety of eddy diffusion profiles ranging from 10 4 to 10 8 cm 2 s 1 below the tropopause. A knowledge of the internal planetary heat fluxes by way of free convection theory give estimates of K(z) for the deep atmosphere of 10 7 to 10 9 cm 2 s 1 (Flasar and Gierasch, 1977; Prinn and Owen, 1976; and Stone 1976). We have modelled the bottom segment K(z) values to be within the constaints thus predicted, but found little impact on any of our results as chemistry dominates over dynamics in the lower troposphere and so we nominally choose K bottom ref = 10 8 cm 2 s 1 as our standard reference in this region. In the middle segment we adopt a standard reference value in this region of K middle ref = 10 7 cm 2 s 1. However, this represents a region where convection is also much more dominant than chemistry and so we explore the range of K(z) from 10 4 to 10 8 cm 2 s 1 as shown in Figure 2. Results and Discussion In the previous section, we have constructed a standard reference eddy diffusion profile which, together with the temperature profile in Figure 1 and chemistry given by reaction (1), over a large pressure range from the bottom of the Jovian troposphere up to

13 13 its mesosphere define a standard reference model atmosphere for the species CH 3 D, H 2, He, CH 4, and HD. We now report on the effects of varying important parameters within their error bars over this pressure regime on the standard reference atmosphere and discuss how this impacts the CH 3 D mixing ratio, CH 3 D fractionation, the ([D]/[H]) CH4 and D/H ratios and compare with the various CH 3 D and HD observations mentioned previously. Figure 3 shows the chemical time constant, τ chem, for CH 3 D. Also shown are the eddy diffusion time constants, τ K = L 2 /K(z) for the various eddy diffusion profiles over the pressure range we have considered (cf. Figure 2) and where the mixing scale length L = H. τ K is a measure of time required to mix the CH 3 D over a scale height. In the deep atmosphere below 300 bar, τ chem < τ K, and so is a region where thermochemical equilibrium obtains making chemistry the dominant process. Above this region where τ chem > τ K, thermochemical equilibrium is not attained and vertical mixing will dominate chemical processes. This is evidenced by examining Figure 4a which shows the calculated CH 3 D mixing ratios for the various cases of K(z) given in Figure 2 over the pressure range considered. At the bottom of the troposphere, all values of the CH 3 D mixing ratio are the same as chemistry dominates there and the effects of changed vertical mixing are only noticed above this region. The exact pressure level at which τ chem τ K varies depending on how vigorous the vertical mixing is. A lower value of K(z) positions τ chem τ K at somewhat lower pressures in the deep troposphere compared to higher values of K(z). We found that changing the CH 3 D mixing ratio at the top boundary within the

14 14 range of values predicted by Parkinson et al. (2000) made little difference to the CH 3 D profiles in the height regime of this figure which indicates that little CH 3 D gets down from the thermosphere of Jupiter into the troposphere. Also, an adjustment of K middle ref from the 20 bar level suggested by Landry et al. (1991) downward to 40 bar had no impact on the CH 3 D mixing ratio profiles. This is because these pressure regions are sufficiently high in the troposphere, well above the region where τ chem < τ K, such that the vertical mixing is completely dominant and so the location of this point in this region is not very important. What is important is the value of K(z) in the middle segment, not the exact location of its lower boundary, and hence, we arbitrarily use K middle ref with the lower boundary at 20 bar for all subsequent calculations. Figure 4b shows the range of CH 3 D mixing ratios for the standard reference atmosphere obtained by changing the reaction rate coefficient, k r, within the published error bars of Lécluse et al. (1996). It is seen that this does not produce large differences in the CH 3 D mixing ratio results either and so we use k r = e ( /T ) in the remainder of our calculations. Figure 5 shows the isotopic enrichment factor, f(z). In panel a, which is from the upper troposphere to the mesosphere, we see that f(z) is constant in the upper troposphere for each of the model atmospheres shown in Figure 4a and approaches the mesospheric value of Panel b shows that at the bottom of the lower troposphere, where thermochemical equilibrium obtains, f(z) approaches unity for all model atmospheres considered, and above this, f(z) > 1, since the atmosphere is not in thermochemical equilibrium. This figure illustrates the impact on the values of f(z)

15 15 between 100 and 10 2 bar for the range of values for K(z) considered. α(t ) is shown for comparative purposes and we note that f(z) α(t ), as it should in the deep troposphere. We find that f(z) varies from 1.15 to about 1.85 as K(z) varies from 10 4 to 10 8 cm 2 s 1 in this region with our standard model corresponding to a reference value of f(z) = Our reference value is comparable to Fegley and Prinn (1988) and Lécluse et al. (1996). The standard reference CH 3 D mixing ratio shown in Figure 4a in the pressure region 1 mbar to about 30 bar is very nearly constant, viz., about , which very similar to the value reported by Encrenaz et al., (1996). Utilising this value we are able to calculate a standard reference ([D]/[H]) CH4 ratio in this pressure regime, which if divided by our standard reference f(z) give us back our original ([D]/[H]) H2 ratio of about , as it should. The upper and lower error bounds are obtained by using the calculated CH 3 D mixing ratio based on K(z) equal to 10 4 and 10 8 cm 2 s 1, respectively. Also, we incorporate the error bars inherent in the Galileo GPMS CH 4 and HD mixing ratio measurements (Niemann et al., 1996; Mahaffy et al., 1998) and f(z). Obviously D/H ratios obtained by CH 3 D spectroscopic measurements must be normalised to the standard reference values for the CH 4 mixing ratio and f(z) used in this paper in order to allow a proper comparison with our calculated D/H value. Hence, these values are converted back to the corresponding ([D]/[H]) CH4 ratio, normalised to the Galileo GPMS CH 4 mixing ratio measurement (Niemann et al., 1996), and then converted back to the ([D]/[H]) H2 ratioby dividing by our standard f(z) value. D/H ratios obtained by HD measurements do not require such normalisation and are left as reported, although

16 16 the corresponding ([D]/[H]) CH4 ratio is calculated for comparative purposes. All of these results are summarised in Table and the D/H ratios are shown in Figure 6. Our standard reference ([D]/[H]) CH4 ratio value is consistent with the CH 3 D ISO observation (Encrenaz et al., 1996) and the Gautier and Owen (1989) CH 3 D observation in the pressure region shown, as well as with the HD and CH 3 D ISO measurement of Lellouch et al., (2001), but not so well with the CH 3 D observations of Carlson et al. (1993) and Beer and Taylor (1973). As for the Galileo probe D/H ratio, our standard reference D/H ratio derived from the CH 3 D modelling is also compatible with the most recent estimates of the canonical protosolar D/H value (Geiss and Gloeckler, 1998; Linsky, 1998; Vidal-Madjar et al., 1998; Reeves, 1998). This makes our model self-consistent in the sense that both the CH 3 D and HD reservoirs lead to the same D/H ratio when the CH 3 D chemistry and the atmospheric structure are described as given in the model description section. Since the CH 3 D mixing ratio and the isotopic enrichment factor in the 1 to 20 bar region is constant for a given K(z) (cf. Figure 4a and Figure 5) one would expect the normalised ([D]/[H]) CH4 values from the various observations to be relatively constant within error bars. In Figure 6, we see that all measurements A to F, save C, could be within the error bars and thus there might be a canonical value. However, this is may also not be the case. Apart from considering systematic errors in the different data sets, a possible explanation could be found in the tropospheric temporal variation of temperature or in the variable nature of K(z). The modelling of small variations in the temperature profile below 1 bar down to

17 bar by as much as ±10K yielded negligble changes in the CH 3 D number density profile. This is due to: a) the temperature being sufficiently low higher up in the troposphere such that any small temperature variations will still not be sufficient to drive the chemical changes or b) the temperature differential relative to the actual temperature in the lower troposphere is too small to effect a change in the number density. As we know, the CH 3 D mixing ratio is strongly dependent upon K(z) in the region of interest: therefore, temporal or latitudinal variations in K(z) could significantly impact the measured ([D]/[H]) CH4 ratio. Indeed, the K(z) adopted must needs represent complex upward convection and downdraft mixing that occurs in the Jovian atmosphere as evidenced by recent observations by Gierasch et al. (2000), Vincent et al. (2000), Roos-Serote et al. (2000), and Encrenaz et al. (1999). Gierasch et al. (2000) report observations by the Galileo spacecraft of an active storm system with a vertical extent of at least 50 km and 4000 km in length where moist convection similar to large clusters of thunderstorm cells on the Earth is the dominant factor in converting heat flow into kinetic energy and facilitates large-scale motions in the Jovian atmosphere. Using the HST/WFPC2 (Hubble Space Telescope Wide Field and Planetary Camera 2), Vincent et al. (2000) found that observed bright and dark latitudinal bands at low and mid-latitudes are correlated with zonal winds in the troposphere: the bright bands in the F160W images are at latitudes of subsidences and the dark bands at upwellings. Roos-Serote et al. (2000) and Encrenaz et al. (1999) indicate that there are humid cells very close to dry hot spots that appear to be associated with large

18 18 storm systems at about 5 bar which would cause a lot of turbulence between these regions and account in part for the large variation in this pressure region. Flaser (1986) estimates vertical velocities at the tropopause of cm s 1 are required to maintain the observed thermal structure at the tropopause if the radiative relaxation time is about 3 years. However, temporal variations in the temperature field also suggest that dynamical processes may be important on short timescales in establishing structure in the tropopause region. A comparison between Voyager 1 and 2 temperature maps indicate significant changes in the tropopause thermal structure happened over a 4 month period (Hanel et al., 1979) implying much greater vertical velocities. Vincent et al. (2000) observe temporal variations in the meridonial scans over a period extending from May 1994 to March Jupiter s dominant large-scale weather patterns are zonal jets and long-lived ovals which have dimesions on the order of 10,000 km (Ingersoll et al., 2000). The jets flow east and west and have constant speeds of up to 180 m s 1 (Ingersoll et al., 1981; Limaye, 1986; Vasavada et al., 1998). From the Galileo data (Gierasch et al., 2000) it is known that lightning, associated with optically thick, high white clouds clusters that appear suddenly at pressure regions of a few hundred millibars and grow to over 1000 km in diameter in a few days, occurs mainly in the Jovian belts which have a disturbed, chaotic appearance. Below these clusters, which last from about 12 hours to several days, are clouds from about 3 to 6 bar where water is the only condensate (Geirasch et al., 2000; Ingersoll et al., 2000). Ingersoll et al. (2000) argue that the belts have a net rising motion at the cloud base, as lightning implies strong moist convection which

19 19 is supplied by moisture-laden air from the Jovian interior. Hueso and Sánchez-Lavega (2001) present a three-dimensional moist convection cloud model for the giant planets. Their model results show for a typical single storm cell with a horizontal size of 60 km that updrafts with a vertical velocity of 60 m s 1 can obtain. Clearly, our K(z) thus represents an ill-defined average for all these processes. For example, on the earth, out of the planetary boundary layer the mean K(z) is 10 5 cm 2 s 1 in the troposphere (p > 70 mbar at the equator), but the K(z) representing transport within a convective cell can be many orders of magnitude higher. Hence, depending on where and when the Jovian measurements are taken, we may have an average value for K(z) that can vary over the constrained values given above. This is shown as follows: if we consider the vertical velocity to conservatively to be in the range 0.01 < v < 100 cm s 1, and L eff = tropospheric pressure scale height, H 10 6 cm), we can easily obtain K(z) between 10 4 and 10 8 cm 2 s 1 using the relation K(z) = vl eff. Now, if we regard each of the CH 3 D measurements shown in Figure 6 to be correct we can deduce an associated K(z) in the deep troposphere. To do this we can use relation (3), to obtain the isotopic enrichment factor when ([D]/[H]) H2 is set to (GPMS value). Using Figure 7, which shows f(z) as a function of K(z), we can obtain the range of K(z) that corresponds to the observed ([D]/[H]) CH4 ratio. The results are shown in Table 6.2. Thus interpolating the ([D]/[H]) CH4 ratios so far obtained from different regions and/or for different periods of time in terms of eddy diffusion coefficients we obtain a range roughly between 10 4 to 10 8 cm 2 s 1. Given our discussion on possible dynamical phenomena this is one means of explaining the CH 3 D abundance

20 20 variability so far observed since Modern spectro-imaging observations that offer a good spatial and spectral resolutions are the appropriate techniques to test the relationship we propose in this study between the CH3D mixing ratio and the dynamic state of the 1-20 bar region tropospheric layer. From a modelling perspective, we can extend the investigation much lower in the atmosphere, and comment accordingly, but the observations to date are constrained between 1-20 bar. Future observations (HST, NGST [Next Generation Space Telescope], Galileo) connected with a re-analysis of archive data, may reveal accurately the space and temporal variations expected. At anytime a global mapping snapshot of the dynamic state of the deep tropospheric layer could be obtained by spatially resolved spectroscopy through a full scan of the planetary disc versus latitudes and longitudes. In any case our calculations show how ([D]/[H]) CH4 might be used as a diagnostic tracer to constrain K(z) and better understand the dynamics of the atmosphere in this pressure regime. Acknowledgments. C. D. Parkinson wishes to acknowledge that some of this work was supported by the National Aeronautics and Space Administration through the NASA Astrobiology Institute under Cooperative Agreement No. CAN-00-OSS-01 and issued through the Office of Space Science. C. D. Parkinson and L. Ben-Jaffel acknowledge support from CNRS and INSU through the PNP program. J. C. McConnell wishes to thank the NSERC for continuing support. We thank Daniel Gautier for the tropospheric temperature-pressure profile used in this work and Scott Edgington for helpful discussion regarding vertical mixing at the

21 tropopause. 21

22 22 References Allen, M., J.P.Pinto, and Y.L. Yung, Titan: Aerosol photochemistry and variations related to the sunspot cycle, The Astrophys. J., 242, L125 L128, Beer, R., and Taylor, F. W., The Abundance of CH 3 D and the D/H Ratio in Jupiter, Ap. J., 179, , Bjorker, E. L., Larson, H. P., Kunde, V. G., The Gas Composition of Jupiter Derived from 5µm airborne Spectroscopic Observations, Icarus, 66, , Carlson, B. E., Lacis, A. A., and Rossow, W. B., Tropospheric gas composition and cloud structure of the Jovian North Equitorial Band, J. Geophys. Res., 98, , Chamberlain, J. W., and Hunten, D. M., Theory of Planetary Atmospheres, Academic Press, Toronto, Edgington, S. G., Atreya, S. K., Trafton, L. M., Caldwell, J. J., Beebe R. F., Simon, A. A., West, R. A., and Barnet, C., On the Latitude Variation of Ammonia, Acetylene, and Phosphine Altitude Profiles on Jupiter from HST Faint Object Spectrograph Observations, Icarus, 133, , Edgington, S. G., Atreya, S. K., Trafton, L. M., Caldwell, J. J., Beebe R. F., Simon, A. A., and West, R. A., Ammonia and Eddy Mixing Variations in the Upper Tropsphere of Jupiter from HST Faint Object Spectrograph Observations, Icarus, 142, , Encrenaz, Th., et al., First results of ISO-SWS observations of Jupiter, Astron. Astrophys., 315, L397-L400, Encrenaz, Th., Drossart, P., Feuchtgruber, H., Lellouch, E., Bézard, B., Fouchet, T., Atreya, S. K., The atmospheric composition and structure of Jupiter and Saturn from ISO observations: a preliminary review, Planet. Space Sci., 47, , 1999.

23 23 Fegley Jr., B., and Prinn, R. G., The predicted abundances of deuterium-bearing gases in the atmospheres of Jupiter and Saturn, Astrophys. J., 119, , Flasar, F. M., and Gierasch, P. J., Eddy diffusivities within Jupiter, in Proc. Symposium on Planetary Atmospheres, ed. A. Vallance-Jones, Royal Society of Canada, Ottawa, 85 87, Flasar, F. M., Global Dynamics and Thermal Structure of Jupiter s Atmosphere, Icarus 65, , Gautier, D., and Owen, T., The composition of outer planets atmospheres, in Origin and Evolution of Planetary Atmospheres edited by S. K. Atreya, J. B. Pollack, and M. S. Matthews), pp University of Arizona Press, Tucson, Geiss, J., and Gloecker, G., Abundances of Deuterium and Helium-3 in the Protosolar Cloud, Space Sci. Rev., 84, , Gierasch, P.J., Ingersoll, A. P., Banfield, D., Ewald, S. P., Helfenstein, P., Simon-Miller, A., Vasavada, A., Breneman, H. H., Senske, D. A., and the Galileo Imaging Team, Observations of moist convection in Jupiter s atmosphere, Nature, 403, , Gladstone, G. R., M. Allen, and Y. L. Yung, Hydrocarbon photochemistry in the upper atmosphere of Jupiter, Icarus, 119, 1 52, Hanel, R. A., et al., Infrared observations of the Jovian system from Voyager 2, Science, 206, , Hueso, R. and Sánchez-Lavega, A., A Three Dimensional Model of Moist Convection for the Giant Planets: The Jupiter Case, Icarus, 151, , Hunten, D. M., Vertical transport in Atmospheres, in Atmospheres of Earth and the Planets, ed. B. M. McComac, 59 72, D. Reidel, Dordrecht, 1975.

24 24 Hubbard, W. B., and MacFarlane, J. J., Theoretical predictions of deuterium abundances in the Jovian planets, Icarus, 44, , Ingersoll, A. P., et al., Interaction of eddies and mean zonal flow on Jupiter as inferred from Voyager 1 and 2 images, J. Geophys. Res., 86, , Ingersoll, A. P., Gierasch, P.J., Banfield, D., Vasavada, A., and the Galileo Imaging Team, Moist convection as an energy source for the large-scale motions in Jupiter s atmosphere, Nature, 403, , Landry, B., Allen M., and Yung, Y. L., Troposphere-Stratosphere Interactions in a One-Dimensional Model of Jovian Photochemistry, Icarus, 89, , Lécluse, C.,, Robert, F., Gautier, D, and Guiraud M., Deuterium enrichment in giant planets, Planet. Space Sci., 44, , Lécluse, C. and Robert, F., Hydrogen isotope exchange reaction rates: Origin of water in the inner solar system, Geochimica et Cosmochimica Acta, 58, , Lellouch, E., Bézard, B., Fouchet, T., Feuchtgruber, H., Encrenaz, Th., and de Graauw, T., The deuterium abundance in Jupiter and Saturn from ISO-SWS observations, Astron. Astrophys., 370, , Limaye, S., Jupiter: New estimates of the mean zonal flow at the cloud level, Icarus, 65, , Linsky, J. L., Deuterium Abundance in the Local ISM and Possible Spatial Variations, Space Sci. Rev., 84, , Mahaffy, P. R., T. M. Donahue, S. K. Atreya, T. C. Owen, and H. B. Niemann, Galieo probe measurements of D/H and 3 He/ 4 He in Jupiter s Atmosphere, Space Sci. Rev., 84, , 1998.

25 25 Mason, E. A. and Marrero, T. R., The diffusion of atoms and molecules, in Advances in Atomic and Molecular Physics, eds., D. R. Bates and I. Esterman, Vol. 6, pp , Academic Press, New York, Massie, S. T., and Hunten, D. M., Conversion of para and ortho Hydrogen in the Jovian Planets, Icarus, 49, , Niemann, H.B., et al., The Galileo Probe Mass Spectrometer: composition of Jupiter s Atmosphere, Science, 272, , Owen, T. et al., Deuterium in the Outer Solar System: Evidence for Two Distinct Reservoirs, Nature, 320, , Parkinson, C.D., E. Griffioen, J.C. McConnell, L. Ben Jaffel and G.R. Gladstone, D chemistry and emission in the Jovian thermosphere, in preparation for Icarus,, Prinn, R. G. and Owen, T., in Jupiter, ed. T. Gehrels, University of Arizona Press, Tuscon, 319, Reeves, H., Concluding Remarks, Space Sci. Rev., 84, , Seakins, P. W., S. H. Robertson, M. J. Pilling, D. M. Wardlaw, F. L. Nesbitt, R. P. Thorn, W. A. Payne, L. J. Stief, Temperature and isotope dependence of the reaction of methyl radicals with deuterium atoms, J. Phys. Chem., 101, , Sieff, A. et al. Thermal Structure Jupiter s upper atmosphere derived from the Galileo Probe, Science, 276, , Smith, W. H., Schempp, W. V., Simon, J., and Baines, K. H., D/H for Uranus and Neptune, Astrophys. J., 336, , Stone, P. H., Meteorology of the Jovian Atmosphere, in Jupiter, ed. T. Gehrels, University of Arizona Press, Tuscon, , 1976.

26 26 Vasavada, A. R., et al., Galileo imaging of Jupiter s atmosphere: The Great Red Spot, equatorial region, and white ovals, Icarus, 135, , Vervack, R. J., B. R. Sandel, G. R. Gladstone, J. C. McConnell, and C. D. Parkinson, Jupiter s He 584 ÅDayglow: New results, Icarus, 114, , Vidal-Madjar, A., Ferlet, R., and Lemoine, M., Deuterium Observations in the Galaxy, Space Sci. Rev., 84, , Vincent, M. B., et al., Mapping Jupiter s Latitudinal Bands and Great Red Spot Using HST/WFPC2 Far-Ultraviolet Imaging, Icarus, 143, , von Zahn, U. and D. M. Hunten, The Helium Mass Fraction in Jupiter s Atmosphere, Science, 272, , C. D. Parkinson Division of Geological and Planetary Sciences, California Institute of Technology/Jet Propulsion Laboratory and the NASA Astrobiology Institute, 1200 E. California Blvd., Pasadena, CA, 91125, USA Institut d Astrophysique de Paris-CNRS, 98bis Blvd. Arago, Paris, France. ( cdp@gps.caltech.edu) L. Ben Jaffel, Institut d Astrophysique de Paris-CNRS, 98bis Blvd. Arago, Paris, France. J. C. McConnell, Department of Earth and Atmospheric Science, York University, Petrie Science Bldg., 4700 Keele Street, Toronto, Ontario, Canada, M3J 1P3 Received ; revised ; accepted.

27 27 Figure 1. The standard temperature profile adopted for Jupiter is based on a smoothed version of the results of Sieff et al. (1997) and Lécluse et al. (1996). Figure 2. Eddy diffusion profiles used to calculate isotopic enrichment values and CH 3 D mixing ratios. Figure 3. Chemical and diffusion time constants for the standard temperature profile. The diffusion time constant curves are for the eddy diffusion profiles given in 2. Figure 4. CH 3 D mixing ratio profiles as a function of pressure for (a) various eddy diffusion profiles considered, and (b) upper and lower bounds for reaction rate k r for standard reference case. Figure 5. Isotopic enrichment factors as a function of pressure for various eddy diffusion profiles considered where (a) focuses on the upper atmosphere and (b) focuses on the lower atmosphere. Figure 6. Calculated standard reference and observed D/H ratios. Measurements included with error bars: A, Beer and Taylor, (1973); B, Gautier and Owen, (1986); C, Carlson et al., (1993); D, Encrenaz et al., (1996); E and F, Lellouch et al., (2001); G, Mahaffy et al., (1998). Also shown is the protosolar D/H ratio, H (Geiss and Reeves, 1993) These results have been normalised to our standard reference values for f(z) and the CH 4 mixing ratio (cf. Table ). Figure 7. Calculated isotopic enrichment factor, f(z), as a function of the eddy diffusion coefficient, K(z)

28 28 Table I Observational Data Reference Detection Pressure CH 4 f(z) Reported Normalised Normalised (cf. (bar) mixing ratio (D/H) H2 (D/H) CH4 (D/H) H2 Fig. 6) ( 10 3 ) ( 10 5 ) ( 10 5 ) ( 10 5 ) A CH 3 D B CH 3 D (Voyager) ± ± ±0.8 C CH 3 D (Voyager) ± ± ±0.7 D CH 3 D ISO (Voyager) ± ± ±0.9 E CH 3 D ISO (Galileo) ± ± ±0.7 F HD ISO (Galileo) 2.4± ± ±0.4 G HD GPMS (Galileo) 2.6± ± ±0.7 H Protosolar 2.1± ± ±0.5 I This study (Galileo) ± ±1.1

29 29 Table II Extrapolated K(z) values Reference f(z) required Estimated K(z) as explained in to normalise to during observation Figure 6 GPMS (D/H) H2 value (cm s 1 ) A 0.76 > 10 8 B > 10 8 C 2.2 < 10 4 D > 10 8 E > 10 8