Manuscript No. MS#I08880
September 16, 2005
Dr. Peter Gierasch,
Editor, Icarus
Space Sciences Building
Cornell University
Ithaca, NY 14853-6801
Dear Dr. Gierasch:
In response to your decision for moderate revisions to our paper “Deuterium chemistry
and airglow in the jovian thermosphere” (MS#I08880), we have revised our manuscript
according to the reviewers’ remarks. It has taken a lengthy time to handle all the issues
properly for such a long paper amidst the myriad of other tasks one does as a research
scientist. We thank the referees for their comments and patience. All the points are well
taken, and we have made a number of revisions to the paper detailed below to clarify the
points raised by the referees. We will first reproduce the referee's questions with italic
font, followed by our responses.
I. From the first referee
(a) General comments
1) A central thesis of this paper, that D in Jupiter's thermosphere is produced primarily
by H + HD --> D + H2, is unproven. The models contain both chemical sources of D
and an external source, specified by a flux of D at the upper boundary of the calculations,
but the relative importance of these sources is never discussed. This problem is relatively
easy to fix by presenting calculations for a zero D flux at the boundary. It is essential
that this be done.
We did the calculation for zero D flux: cf.: Parkinson (Thesis, 2002); Parkinson et al.
(1999). The validity of the reaction the reviewer terms “unproven” has been discussed in
detail by Parkinson et al. (1999)
2) The method for calculating rates for reverse reactions from the forward rate and the
equilibrium constant is poorly described. The calculation appears in several places in the
paper but is never adequately discussed. The brief discussion that is included is, at least
techinically, incorrect. The ratio of forward and reverse reactions is equal to the
equilibrium constant for the reaction, which is equal to the exponential of the change in
the Gibbs free energy, not the energy difference between levels. I did a quick check of the
ratio of forward to reverse rate constants for H + HD --> H2 + D. Using
thermodynamic data from the JANAF tables at 1000 K I calculate k_f/k_r = 0.38. The
formula given on page 9 of the manuscript implies k_f/r_r = 0.33. I am not claiming that
my value is better, but the authors need to supply enough information so that the reader
can understand their approach.
It seems to me that the H2 ortho-para ratio might play some role in these thermodynamic
calculations, but it is not mention.
We have included a short discussion describing the method of calculating the rates for the
reverse reactions. Also, for these studies, the pressure region of interest is extensive and
extends from the mesopause to well down in the troposphere to 1000 bar. The lower
boundary (1000 bar) is set to such a high pressure to ensure a depth to where
thermochemical equilibrium obtains. At such depths, effects due to ortho-para H2
conversion should be considered. Massie and Hunten (1982) and Carlson et al. (1993)
treat ortho-para H2 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 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 CH3D in the
troposphere of Jupiter and so we do not consider this effect in our calculations.
3) I realize that the details of the atmospheric model are not essential to the main points
of the paper, but I find some of the choices made for the atmospheric model to be odd.
For example, the temperature profile is very jagged while the actually Galileo ASI
profile, upon which it is based, is smooth. Perhaps the profile used here was digitized at
low resolution from a plot? The actual profile is available through the PDS. Why wasn't
it used? Also, the Sieff et al. results extend only to 1000 km, whereas the profile shown in
Fig 2b. has structure at higher altitudes. Where does this come from?
The Voyager UVS occultation experiment at Jupiter has provided a good estimate of the
CH4 density in the vicinity of the homopause. Results are presented in Yelle et al.
(1996). These results were later confirmed by Drossart et al. (1999) through analysis of
CH4 fluorescent emissions. The Vervack et al. results were based on observations of He
emissions. Since the authors are interested in the CH4 densities, it would seem to make
more sense to use the CH4 constraints rather than trying to extrapolate from the He
results.
We used a reasonably accurate digitized profile obtained from the Sieff et al. (1997)
paper. We have extended the result to include higher altitudes and are representative of
exospheric temperatures.
(b) Specific comments
1) page 2, line 7. Change "... depend on the uncertainty in some of the reaction rate
constants used." to " ... depend on the uncertain values of some reaction rate
constants."
This has been fixed.
2) First paragraph, first sentence. I believe that the D/H ratio was first estimated by
Trauger et al. by measuring the pressure induced HD line.
This reference has now been included.
3) page 5, continued paragraph at the top of the page. The main result from the Yelle et
al. reanalysis of the Voyager UVS data was that the temperature gradient in the lower
atmosphere was 3 K/km, not 1 K/km. The Galileo measurements verified this. The
work by Yelle et al. and Sieff et al. shows that the results of Festou et al. (1981), used
by Romani et al. (1993) are wrong. Why is it used here? By the way, the Romani et al.
temperature profile is cooler than the Sieff et al. profile, not warmer, as stated in the
text. This is a source of some confusion since it appears that the calculations based on
the Romani et al. and Sieff et al. profile is consistent with Romani et al. profile being
warmer.
We use the Sieff et al. result as our standard case. As the reviewer points out, the Sieff et
al. result shows that the results of Festou et al. (1981), used by Romani et al. (1993) are
wrong. However, we use the Romani (1996) profile since we are interested only in
considering an atmosphere profile that is cooler than that obtained by the Galileo Probe
measurement (Sieff et al., 1997) for comparative purposes, and in no way regard the
Romani temperature profile as canonical.
4) page 5, lines 6-7. Yelle et al. (1996) did not incorrectly predict the temperature of the
upper thermosphere. In fact, they did not predict the temperature of the upper
thermosphere at all, but rather constructed a model using the temperatures inferred by
Marten et al. (1994). The Marten et al. results are within the range of temperatures
derived by Sieff et al., so in fact there is no disagreement. Please correct this sentence.
This has been fixed.
5) Page 5, equation 1. The relationship between rms velocity is well known and not
essential to this paper. You don't need this equation.
This equation not absolutely essential, but we are leaving it in for pedagogical reasons.
6) Page 7, continued paragraph. I can't follow this paragraph. The authors state "This
experiment yielded information on the total Lyman-alpha flux, the central Lyman alpha
flux, the blue wing flux at D Lyman-alpha, and the slope of the D Lyman-alpha at the
same location[,] which helps us [to] overcome one previously unresolvable aspect of
solar UV irradiance variation: the shape of [the] H Lyman alpha solar line and its
variability with solar activity." This thought is never finished; we are never told how the
problem has been solved. (The square brackets in the quote are mine and indicate where
grammatical corrections are necessary.)
We have made the grammatical changes. It isn't our purpose to explain the details of the
H Ly-alpha solar line and its variability with solar activity here; moreover, the references
are thorough and our assertion easily verified.
7) Page 8, equation 7. See general comment 2 above.
We have noted this.
8) Page 8, line 24. There are papers by Abgrall et al. in the peer-reviewed literature for
the energy levels of H2 and HD. The reference to private communication from Mandy is
inappropriate. If there are differences between those values and the Abgrall et al.
results, they should be discussed in this paper. The references to private communication
for H2 energy levels occur throughout the text and should be corrected everywhere. The
Abgrall et al. calculations are used by most workers in the field.
This is now fixed and we are employing the correct reference now.
9) Page 9, line 9. See general comment 2 above.
We have noted this.
10) Recent work on photochemical modeling of the Jovian atmosphere is discussed in the
review chapter by Moses et al. to be published in the Jupiter book (by Cambridge
University Press) this year. This work should be discussed and compared with the model
presented in this paper.
The chemical model of Moses et al. used in the Jupiter book mentioned is a derivative of
the Caltech/JPL KINETICS model (Allen et al., 1980; Gladstone et al., 1996). We have
discussed/validated our model in relation to this model in detail in section 2.4
Photochemical Model and in Appendix A.
11) Page 9, line 10-11. A reference is not needed for the principle of detailed balancing.
We have noted the reviewer’s remarks but have left the reference in for other readers.
12) Page 9, line 10. ""... where we have simplified the distribution of states by taking the
energy difference between the v=0 and J=0 states supplied by M. Mandy (private
communication, 1999)" This sentence should be rewritten. I am not sure what
"simplified the distribution of states" means in this context. It is not clear what energy
difference is referred to. Do you mean the energy difference between H2(v=0,J=0) and
HD(v=0,J=0)? Isn't the difference between HD and H2 rotational levels important?
Also, as mentioned above, the ratio of forward and reverse reactions is equal to the
equilibrium constant, which is equal to the exponential of the difference in the Gibbs free
energy, not the energy difference between levels. This must be corrected. See general
comment 2 above.
We have changed this now in the text to reflect the reviewers comments.
13) Page 9, line 21. Private communication references to a co-author are not allowed,
are they? After all, since Yung is a co-author on this paper, his contribution is already
acknowledged.
We overlooked this change after Y.L. Yung became a co-author on the paper. We have
changed this now and removed the private communication.
14) Page 10, line 14. Since you do not include ion chemistry and the associated
production of H atoms in your model, it seems to me that there is no place for an
assumption about the number of H atoms produced by each absorbed photon. In any
case, most of the H is produced in the aurora, not by solar EUV, which is why the H flux
required to match the column density inferred from observations is a factor of 4 larger
than flux produced by the ionosphere.
The reviewer is correct that we do not include ion chemistry and the associated
production of H atoms in our model as described. However, we feel this is a reasonable
assumption consistent with Gladstone et al. (1996) and have left the discussion as is.
15) Page 10, line 25. It seems to me that your assumption that the D flux at the top of the
atmosphere is half the H flux time the D fraction is questionable. There are two issues.
First, as mentioned above, most of the H is produced in the auroral zones. D will be
produced there as well, but the meridional transport of H and D may differ because of
the difference in their molecular weight. D should diffuse downward faster than H and
the amount of D that is transported from the auroral zones to low latitudes may be
smaller than for H. The second issue is that the HD bond differs from the H2 bond and
the HD ionization rate may differ from the H2 ionization rate. It should be possible to
say something about this by examining electron impact ionization cross sections for H2
and HD. Both these issues require much more extensive discussion than currently
contained in the paper.
We note the reviewer’s points and don’t disagree, but feel this is a reasonable assumption
consistent with Gladstone et al. (1996) and Parkinson (Ph.D. thesis, 2002). Moreover, we
clearly state that we are considering this to be a standard reference value in the text.
Examination of Figure 6 shows that the effect of varying the input H flux (and hence D
flux) to be relatively small compared to the effects of varying the vibrational temperature
profile and so our choice appears adequate for this work. Future work to be done would
include a more detailed analysis discussing the implication of downward fluxes that are
of auroral origin: then one might estimate separately the H and D fluxes taking into
account transport effect on constituents from auroral to equatorial regions.
16) Paragraph that starts on the bottom of page 11. I can't follow this. Do the radiative
transfer calculations solve the coupled or uncoupled problem?
As we say in the text, “Parkinson et al. (1999) applied a resonance scattering model to the
D + H problem … to solve the coupled equations … and for the uncoupled H and D
cases for subsolar conditions...” i.e. we solve the D + H coupled problem as our main
case and uncoupled H and D for comparative purposes for a limited set of conditions.
17) page 12, line 7. A minor point, photons are not "completely absorbed" at tau=1, but
only by a factor of e^-1. The sentence should be rewritten.
We have rewritten the sentence as the reviewer has suggested.
18) Page 12, line 29. You ask the reader to compare Figs 2b and 2c to see the effect of
vibrational temperature, when I think that you mean to ask us to compare Figs. 2a and
2b.
This correction has been made.
19) Figure 2e shows that the D density in the models using the Romani et al. temperature
profile are larger than the density in the models using the Sieff et al. profile, but the
Romani et al. temperature profile is cooler than the Sieff et al profile! Something is
seriously wrong here, but I can't tell what.
The reviewer is absolutely correct: this figure was inexplicably wrong and we have
corrected this by redoing the figure and changing the text in the manuscript accordingly
to reflect the correct situation.
20) Page 15. The importance of auroral production of H was first pointed out by Yung
and Strobel (1980) and Broadfoot et al. (1981). These should be used in place of the
Waite et al. (1983) and McConnell et al. (1982) references.
We have included these references along with the other references.
21) Page 16, line 34. Another private reference to a co-author.
We have removed the “private communication” and put in the proper reference.
22) Page 16, line 34. Some discussion of the optical depth at line center for D Lyman
alpha would be useful here.
We have discussed optical depth of the D line not only in the line core, but also in the
wings, which is essential to the interpretation of our results. Upon re-examination of this
paragraph in entirety as well as the one following in relation to the figures, we have left
the text as is since the discussion seems complete to the authors.
23) Table VI. Reaction R1. H2 is not dissociated by Lyman alpha.
We have fixed this typogrpaphical error.
24) Fig 1. The caption is misleading because the Romani et al. temperature curve only
uses the temperatures inferred from the Voyager IRIS observations in the stratosphere,
outside the range in the plot.
We made a note to the reader to see the text for details, and explain in the section
“Description of Models” the choices we made in chosing an exospheric temperature to
extend the model for our calculations.
II. From the second referee
(a) General comments
This paper presents parametric studies of deuterium chemistry in Jupiter’s upper
atmosphere, and predicts the brightness of the coupled H and D Lyman alpha dayglow.
The effects of H2 vibrational temperature, the homopause eddy diffusion coefficient, and
certain reaction rate coefficients on the profiles of D-containing species are investigated.
This work is a natural extension of earlier work in this area by Parkinson, and is worthy
of publication in Icarus, after minor modifications. The results indicate that the D column
abundance is primarily sensitive to H2 vibrational temperature, which is reasonable. The
major problem with this work is that H2(v) is not actually calculated, but just assumed to
be various multiples of the ambient atmospheric temperature (e.g., Fig. 4 of Majeed et
al., 1991 shows why this is a poor assumption). So, this work will eventually need to be
redone, perhaps with KINETICS accounting for each vibrational level of H2 as separate
species. Meanwhile, this paper at least gives us an idea of what to expect. The paper is
brief enough, but I think that there are far too many figures. These could be combined
(e.g., using colored lines) or just removed (e.g., Fig. 4d and Fig. 7 are not very
informative).
The reviewer claims a major problem with this work is that H2 vibrational temperature
profile is not actually calculated, but just assumed to be various multiples of the ambient
atmospheric temperature. Utilising different T profiles provides a diagnostic to test the
sensitivity of each of the results to the kinetic temperature profile chosen, since both
kinetic and vibrational temperature profiles will be affected. Cravens (1987) and Majeed
et al. (1991) estimate H2 effective vibrational temperatures between 1300--4000K
depending on altitude and vibrational level. So, we have used the H2 effective
vibrational temperatures of Cravens (1987) and Majeed et al. (1991) as a guide to
generate an arbitrary simple, representative class of T_v profiles in order to obtain
vibrationally excited HD profiles. We have considered T_v between 1 and 4 times
kinetic, with T_v = 3T as our standard reference. Also, contributions to the vibrational
temperature from levels v>1 were found to be relatively small and we did not include
them. However, this work could eventually be redone including higher vibrational levels
for completeness, perhaps with the Caltech/JPL KINETICS code accounting for each
vibrational level of H2 as separate species as suggested along with possibly using an
observationally based H2 vibrational temperature profile such as Gustin et al. (2004).
The latest CH3 recombination rate just measured by Cody et al. (2004), newer data of
Lemaire for Lyman alpha profiles from SOHO/SUMER available, and sensitivity studies
regarding the revised H column and hot' H column estimates of Gladstone (2004) would
also be items to include in future work.
We appreciate the reviewer's concern with the number of figures, but since these figures
are the result of thorough sensitivity studies (very similar to the long paper the reviewer
had in his 1996 Icarus paper) we feel they should be included. The authors have no issue
if the editor would like to bundle some of the figures together at his discretion.
(b) Specific comments
p.3, l.7 – “... excited HD is also ...” -> “... excited HD, is also ...”
This typographical error has been fixed.
p.3, l.18 – 3.7E17 cm-2 is a fairly large column of H. Check out Gladstone et al. (PSS,
52, 415, 2004), who get good fits to Galileo UVS Lyman alpha measurements with a
column of 1.0E17 cm-2 ambient and 1.0E14 cm-2 (0.1%) ‘hot’ hydrogen.
We checked the reference provided by the reviewer and express our thanks. The reasons
we chose the column H we did is explained in the text. An investigation into the effect of
the location of the `hot' region versus altitude was carried out and showed little impact on
the results. We proceeded by initially locating the `hot' region at the top of the
atmosphere and then adjust it downwards until the bottom of the region was located at
about 1900 km. Below this altitude, the `hot' region is presumed to not be sustainable
due to a high collision rate with the `colder' background gas (cf.~Appendix B). In
addition, halving or doubling the column amount of `hot' H changed the D~lyman alpha
brightnesss by less than 7%. This was expected since most of the deuterium is much
lower in the atmosphere than where the `hot' H resides. Since the inclusion of a `hot'
region is seen to minimally affect all calculations, no figures have been included
regarding this parameter.
p.3, l.29 – Hasn’t the CH3 recombination rate just been measured by Cody?
Yes, this recombination rate was recently measured by Cody, but at the time of this study
was unavailable. However, we discuss the implication of the inclusion of this rate in the
“Conclusions” in the fourth paragraph.
p.4, l.14 – “4.4E6 cm2 s-1” -> “4.4E11 cm-2 s-1”
We have fixed this typographical error.
p.4, l.15 – The Tex=1100K is pretty old – why didn’t you use 940K, which is more
consistent with Seiff et al.?
We found that using either value for T_ex made little difference to our results given that
most of the D and HD are well below the exosphere and hence just picked one of them
as our standard reference T_ex.
p.5, 2nd paragraph – Please explain the basis for assuming Tv = nT, which is used
throughout the paper
The basis for assuming T_v=nT is give above in the Reviewer 2 General Comments.
Temperature profiles are so few in number that choosing several of these representations
as we have done is just to explore parameter space…none are canonical!
p.5, l.38 – “Figure 1c” – the manuscript doesn’t have this figure!
The figure is now included!
p.6, l.20 – “analytic approximation” -> “empirical approximation” BTW, the wings of
the solar Lyman alpha line are much better fit if equal and offset Lorentzians (using the
same offset and dispersions) of equal weight are added to the Gaussians
We have fixed this typographical error; comment noted.
p.7, l.6 – FYI, there are a lot newer data from SOHO/SUMER available. Have a look at
the Lyman alpha profiles of Lemaire at
http://medoc-ias.u-psud.fr/instruments/SUMER/sumer-results.html
Regarding available newer data from SOHO/SUMER: see Reviewer 2 General
Comments
p.7, l.28 – typo “soley” -> “solely”
We have fixed this typographical error
p.7, l.39 – “thermoneutral” – is this a real word?
"Thermoneutral" IS a real word: check Google!
p.8, l.15 – “... are small and P_HD^1 = L_HD^1.” -> “... are small.” (the assumption
that P_HD^1 = L_HD^1 is one of steady state, not about the size of R2 and R3)
We have adjusted the wording to reflect the reviewer’s remarks.
p.8, l.24 – “... (private communication, M. Mandy, 1999).” It doesn’t seem necessary for
a reference here, as this is pretty straightforward.
We have removed private communication and gave proper reference.
p.9, 16 – R7 should have + M added to each side (for consistency with the usage in the
rest of the paper)
We have included "+ M" in equation for consistency for R7 everywhere in paper.
p.9, l.21 – “Yung (private communication, 2001) pointed out that Parkinson et al. (1999)
had omitted ...” -> “Parkinson et al. (1999) omitted ...” (I don’t think a reference to a
private communication from a co-author is needed)
We have fixed this error.
p.10, l.16 – “... (eg. Gladstone ...” -> “... (e.g., Gladstone ...”
We have fixed this typographical error.
p.10, l.17 – “requisite” -> “assumed” (again, I think this column is 4x larger than
needed to match the dayglow in the non-bulge region)
We have fixed this phrase and the comment is noted; please see Reviewer 2 comment p.3
1.18 regarding the column amount choice.
p.11, l.7 – I don’t think there should be an alpha_i term in (15)
We have fixed this typographical error.
p.11, l.10 – The dT/dz term is missing a factor of ni in (16) (actually, this equation would
look better written as (15) is now, just replacing K with D and Hav with Hi)
We have implemented this as suggested by the reviewer.
p.11, l.14 – “... and K = K(z) are ...” -> “... and K are ...” (it’s not like T and Di and
practically everything else aren’t also functions of z, so no need to single out K)
We have removed the z dependence for consistency as suggested by the reviewer.
p.11, l.21 – “C4Hy” ? why not C4Hx?
We have fixed this typographical error.
p.11, l.26 – “Parkinson et al. (1999) have applied ...” -> Parkinson et al. (1999) applied
...”
We have fixed this typographical error.
p.12, l.10 – “We also note ... in altitude.” This sentence is awkward and can be deleted.
We did not delete this sentence as suggested by the reviewer since this obvious point
regarding details of the radiative transfer model might not be transparent to the less
informed reader.
p.13, l.1, - “... an slight ...” -> “...a slight ...”
We have fixed this typographical error.
p.13, l.6 – What is the basis for the approximation Tv=3T?
Basis for approximation T_v=3T; see General Comments. Note: this is an arbitrary
choice for a standard reference.
p.13, l.30 – So why does D decrease with Tv at higher z?
Any produced higher than 600km also flows downward contributing to the decrease in D
we see above this altitude.
p.14, l.26 – Isn’t a likelier explanation (for decreasing CH3 at lower altitudes, with
increasing Kh) just that there’s less photolysis since tau_CH4=1 moves higher up?
We agree with the reviewer and have included this comment.
p.14, l.33 – Figure 4d could be removed, and this result just stated.
We have noted the reviewer’s preference, but have chosen to leave the figure in for
comparative purposes.
p.15, l.3 – “... for fixed Kh.” -> “... independent of Kh.”
We have fixed this typographical error.
p.15, l.19 – “... the H column density ...” -> “... the H flux ...”
We have changed this as recommended by the reviewer.
p.15, l.25 – “We now look ... CH3D profiles.” This paragraph could be deleted, along
with Figure 7, as there is no useful information here.
Since we are doing a thorough sensitivity study, we have chosen to leave figure and
paragraph included.
p.16, l.3 – “... the ‘hot’ region ...” -> “... the ‘hot’ H region ...”
We have fixed this typographical error.
p.16, l.22 – FYI, there are very good long-slit echelle spectra of the H Lyman alpha
profile across the disk (from HST-STIS) that show no hint of D Lyman alpha, probably to
a couple of 100R. BTW, why exactly is D Lyman alpha limb darkened – the column
densities are not that large? I would think that both D and H Lyman alpha would show a
peak just off the limb (i.e., at SZA = 90 degrees)
FYI noted. See paper text for discussion regarding D Ly-alpha limb darkening (cf.
Results and Discussion, last four paragraphs.
p.17, l.14 – “... are (a) 770 R ...” -> “... are (a) 760 R ...” (right?)
We have fixed this typographical error.
p.18, l.17 – “The authors ...” -> “We ...”
We have included this change as suggested by the reviewer.
p.18, l.20 – “negligble” -> “negligible”
We have fixed this spelling error.
p.19, l.23 – FYI, Cody et al. (JGR-E, 108, E11, 5119, 2003) have measured 2CH3 + M > C2H6 + M at T=155K
We have changed the text to include mention of Cody et al. (2003) and the implications
of including his recombination rate.
p.25 – Seiff and Ollivier are misspelled
We have fixed these spelling errors.
Table I – 1 sdu = 0.0116 A (not 0.116A)
We have made this correction.
Table IX – first reaction is messed up; also, columns of the rate coefficients used here
and in Gladstone et al. might be added (otherwise the table is not very useful)
Other than the missing “M” on the RHS of the equation, the authors aren’t clear on what
the problem is with the first reaction in this table. This table is very useful for the
discussion that follows in “Appendix A” which makes clear the differences/similarities as
well as performing a model validation between the model used in this work with the
Caltech/JPL KINETICS model which has been published and used elsewhere
extensively.
Fig. 2d – C2H5D seems to be missing
In the figure caption we say the figure is similar to figure c, not the same. The impact on
C2H5D was seen to be negligible as can be seen also with CH3D that is included in the
figure.
Fig. 4a – What are the arrows and heights marking the right axis for?
The arrows on the graph show the range for tau_CH4=1 level for the CH4 profiles
included. This is now explained in the text.
Fig. 4b – What CH4 profile is presented here (and why)?
This is the CH4 profile for the case Tv = 3T, as indicated in the caption. This is now
explained in the text.
Fig. 4c – ditto
See the previous item regarding Figure 4b. This is now explained in the text.
Fig. 4d - What are the arrows and heights marking the right axis for?
See the remarks associated with those above for Figure 4a. This is now explained in the
text.
Fig. 5 – Caption ends suddenly
Figure 5's comment doesn't end suddenly: it is split between two panels 5a and 5b.
Fig. 6 – Caption “... varying the H column density by changing input H flux, ...” ->
“varying input H flux, ...”
The figure caption has been changed to reflect the reviewer’s comments.
Fig. 8 – Does the diffusion time constant include both Di and K? You might show the H
chemical time constant as well
The diffusion time constant includes both D_i and K as we indicate in the text. We didn’t
think to include the H chemical time constant since D and HD were mainly relevant to
our discussion.