Greenhouse gas effects on the solar cycle response of water vapour and noctilucent clouds

Abstract. The responses of water vapour (H2O) and noctilucent clouds (NLCs) to the
solar cycle are studied using the Leibniz Institute for Middle Atmosphere
(LIMA) model and the Mesospheric Ice Microphysics And tranSport (MIMAS)
model. NLCs are sensitive to the solar cycle because their formation depends
on background temperature and the H2O concentration. The solar cycle
affects the H2O concentration in the upper mesosphere mainly in two
ways: directly through the photolysis and, at the time and place of NLC
formation, indirectly through temperature changes. We found that H2O
concentration correlates positively with the temperature changes due to the
solar cycle at altitudes above about 82 km, where NLCs form. The photolysis
effect leads to an anti-correlation of H2O concentration and solar
Lyman-α radiation, which gets even more pronounced at altitudes
below ∼ 83 km when NLCs are present. We studied the H2O
response to Lyman-α variability for the period 1992 to 2018,
including the two most recent solar cycles. The amplitude of Lyman-α
variation decreased by about 40 % in the period 2005 to 2018 compared to
the preceding solar cycle, resulting in a lower H2O response in the
late period. We investigated the effect of increasing greenhouse gases
(GHGs) on the H2O response throughout the solar cycle by performing
model runs with and without increases in carbon dioxide (CO2) and
methane (CH4). The increase of methane and carbon dioxide amplifies the
response of water vapour to the solar variability. Applying the geometry of
satellite observations, we find a missing response when averaging over
altitudes of 80 to 85 km, where H2O has a positive response and a negative
response (depending on altitude), which largely cancel each other out. One main finding
is that, during NLCs, the solar cycle response of H2O strongly depends on
altitude.


1. Lines 14-16: Potentially solar activity can also affect the circulation and transport of the H2O and CH4 from the troposphere.
Yes, you are right. However, in our MIMAS model, we use atmospheric dynamics for all runs corresponding to 1976, so the effects of solar activity on the circulation and transport of H2O and CH4 from the troposphere are not included. Since the effect of the solar cycle on dynamics in the atmosphere is still under debate, we have not included changes in the dynamics to have a better separation of the influences of temperature and Lyman alpha (see the reply to the comment 20).
The decrease in Lyman alpha variability in the late phase is given in absolute values. We have deleted the sentence that partially duplicates lines 22-24 (see the response to reviewer 1).
3. Lines 33-34: Little explanation of the mechanism would be nice.
These lines are deleted (see the response to reviewer 1). The mechanism is described in more detail in section 3.3. 4. Line 50: Which trends? Any trends?
We mean the trends in background temperatures and H2O concentrations. We have added this to line 47 (revised manuscript).

"NLCs have been proposed as indicators of trends in background temperature and H2O
concentrations" 5. Lines 72-78: Has it already been performed and described somewhere, or it is planned for the current manuscript? These lines are confusing because after that they mentioned the goals of the paper. Some rewriting would be nice to distinguish between the introduction and motivation.
We have changed the sentence to make it clearer (Lines 70-72 of the revised manuscript).
The model runs without microphysics were performed in the current study.
"Therefore, for this study, simulations are performed with and without microphysics using the same background conditions, resulting in a H2O profile with and without NLC".  Photolysis is taken into account in MIMAS and causes changes in the H2O concentration.
The increase in CH4 affects the H2O input at the lower boundary of the model (see also Lübken et al., 2018 (section 2)). The sentence has been modified for clarity (lines 108-114 of the revised manuscript).
"Below the MIMAS lower boundary two effects determine the mixing ratio of H2O in the mesosphere: (i) transport of H2O from the troposphere and (ii) oxidation of methane (CH4). The oxidation of each CH4 molecule produces two H2O molecules. Methane is nearly completely converted to H2O in the stratosphere by photochemical processes (e.g., Lübken et al., 2018). MIMAS assumes that transport from the troposphere is constant. The increase in H2O is primarily through (ii) i.e. due to the increase in CH4 concentration (Lübken et al., 2018)".
10. Line 118: 40 million is enough? Any justification? What is the source and properties of the dust particles?
We believe that 40 million is enough. We investigated this in earlier studies in more detail. We added a description of the source of dust particles and included three references for more details (lines 118-120) "Dust particles are formed from meteors evaporating in the atmosphere (for more details, see Berger and von Zahn, 2002;von Zahn andBerger, 2003, Killiani, 2014)". Yes, you are right. We have added some text to this in lines 161-163 (revised manuscript).

"Certainly, at low and middle latitudes, without NLCs one can detect anticorrelation. For example, in H2O satellite data averaged over the tropics (30• N-30• S), an anticorrelation is observed for the "late" period (Karagodin-Doyennel et al. 2021)"
14. Lines 161-162: Do the authors mean temperature effect from solar variability or from GHG or from GWD via circulation?
Our sentence has a general statement without clarification of the temperature variability source. In our numerical experiments, we consider only the variability of temperature due to solar variability and the GHG effect. "The results in Figure 4 illustrate the freeze-drying effect described above and also indicate that the effects of NLC on H2O are not present below ~79 km and above ~97 km. This is the novelty of the results in Figure 4". 16. Line 253: to atmospheric absorption by which species. Should be another maximum lower down due to ozone absorption.
We mean atmospheric absorption by molecular oxygen and water vapour. We added this in line 259 (revised manuscript).
"Temperature differences decrease as altitude decreases because the intensity of solar radiation decreases due to atmospheric absorption by molecular oxygen and water vapour" 17. Line 275: 'increase' or 'production' is missing Line 282 (revised version) modified by including your suggestion.  20. Conclusions: It would be nice to discuss limitations (e.g., fixed GWD) and uncertainties in the applied models. I would also expect some discussion about the statistical significance of the results related to the internal variability of the model based on first principles.
We have studied in detail the model setup and the internal variability (see also answer to Q10). Please find in the figure below an example of the mean and the standard deviation of the July mean data. The dataset includes per altitude 120 x 6 x 4 x 31 = 89280 data points. So the standard error of the mean is about 1/300 of the standard deviation. In conclusion, given the current model combination, we believe that uncertainties in the internal variability (or to low sampling thereof) of the model do not affect our conclusions. We discussed the limitation of using constant dynamics and GWD in lines 441-443 (revised manuscript).
"It should be noted that our results have limitations as they use constant dynamics for all years. We are looking forward to a new gravity wave resolving model to investigate the effects on changing dynamics due to changing GHGs and solar activity".
All typos were corrected .
Other changes are related to the recommendations and demands of other referee.
Thank you for taking the time to review our manuscript.
With respect.