Study of the main processes driving atmospheric CH 4 variability in a rural Spanish region

Atmospheric concentrations of the two main greenhouse gases (GHGs), carbon dioxide (CO 2 ) and methane (CH 4 ), are continuously measured since November 2012 at the Spanish rural station of Gredos (GIC3), within the climate network ClimaDat, together with atmospheric radon ( 222 Rn) tracer and meteorological parameters. The atmospheric variability of CH 4 concentrations measured from 2013 to 2015 at GIC3 has been analyzed in this study. It is interpreted in relation to the variability of measured 222 Rn concentrations, modelled 222 Rn fluxes and modelled heights of the planetary boundary layer (PBLH) in the same period. In addition, nocturnal fluxes of CH 4 were estimated using two methods: the Radon Tracer Method (RTM) and one based on the EDGARv4.2 bottom-up emission inventory. Both previous methods have been applied using the same footprints, calculated with the atmospheric transport model FLEXPARTv6.2. Results show that daily and seasonal changes in atmospheric concentrations of 222 Rn (and the corresponding fluxes) can help to understand the atmospheric CH 4 variability. On daily basis, the variation in the PBLH mainly drives changes in 222 Rn and CH 4 concentrations while, on monthly basis, their atmospheric variability seems to depend on changes in their emissions. The median value of RTM based methane fluxes (FR_CH 4 ) is 0.17 mg CH 4  m −2  h −1 with an absolute deviation of 0.08 mg CH 4  m −2  h −1 . Median methane fluxes based on bottom-up inventory (FE_CH 4 ) is of 0.32 mg CH 4  m −2  h −1 with an absolute deviation of 0.06 mg CH 4  m −2  h −1 . Monthly FR_CH 4 flux shows a seasonality which is not observed in the monthly FE_CH 4 flux. During January–May FR_CH 4 fluxes present a median value of 0.08 mg CH 4  m −2  h −1 with an absolute deviation of 0.05 mg CH 4  m −2  h −1 and a median value of 0.19 mg CH 4  m −2  h −1 with an absolute deviation of 0.06 mg CH 4  m −2  h −1 during June–December. This seasonal doubling of the median methane fluxes calculated by RTM at the GIC3 area seems to be mainly related to the alternate presence of transhumant livestock in the GIC3 area. The results obtained in this study highlight the benefit of applying independent RTM to improve the seasonality of the emission factors from bottom-up inventories.

The paper presents atmospheric data of CH4 mole fractions and Rn222 concentrations observed at a measurement site in central Spain. Surface-atmosphere exchange fluxes of CH4 are estimated based on the radon tracer method, and compared to values from an emission inventory. The topic fits well in the scope of ACP. In general the paper is well written, and I recommend publication after the following concerns have been addressed.
General Comments: The authors found a strong disagreement of Rn based CH4 flux estimates with the values in the EDGAR inventory. Potential reasons for this should be discussed in more detail. What is the contribution in the regional EDGAR CH4 emissions from different source sectors, e.g. enteric fermentation? Which sector seems to be the main cause C1 for the disagreement? Discussing such questions would allow for inventory people to better learn from such observationally based estimates.
Footprint calculation: What was used as the height below which particles are assumed to be influenced by surface fluxes? Ln 210 mentions 300 m, but what was assumed in cases with a nocturnal boundary layer height below 300 m? Particles above the top of the nocturnal boundary layer should not be influenced by surface fluxes. If the method assumes all particles below 300 m to be influenced by surface fluxes, the associated uncertainty in the footprint should be described. Note that usually there is strong wind shear near the top of the nocturnal boundary layer, which worsens a potential error in estimated footprint area. Also it is unclear how exactly the weighting function w(x,t) (Eq. 2) was normalized, and what the exact time limits in the summation in Eq. 2 are. This needs to be clearly described.
Please use an equation to better illustrate the FLEXPART Radon-tracer method derived CH4 fluxes (FR_CH4).
Rather than showing a somewhat hard to read map in Fig 1, why not show the footprint map and a map of the inventory based emissions? That would be better related to the rest of the manuscript.

Specific comments
Ln 90: "flux in this area is of about" I suggest to drop the "of" Ln 124: "The instrument accuracy for CH4 is of 0.36 ppb" I suggest to drop the "of" Ln 143: Is the canopy really below 20 cm? May be this should read "below 20 m"?
Ln 157: Please rephrase the section header, and avoid unreadable terms (i.e. avoid underline characters).
Ln 177: For which time intervals was the correlation between CH4 and Rn assessed, for a single night? This should be stated C2 Ln 231: replace "is" by "of" Ln 242: drop "of" Ln 243: "it is of 30 ppb" drop the "of" Fig. 3 and Fig. 4: it would be useful to show the monthly boxplots also separately for day and night, especially for attributing changes in daily amplitudes; it could well be that low nocturnal PBLH drives the larger amplitude during summer rather than the deeper mixing during daytime as stated in Ln 293. Ln 336: "is of" drop the "of" Ln 336: Looking at the red circles in Fig. 9 it seems that the mean should be much lower, somewhere around 0.1 mg CH4 m-2 h-1. Fig. 9: the grey shaded rectangles seem to be at the wrong position. In the figure caption, e.g. week 21-27 June 2014 is mentioned, while the rectangle seems to be at around mid-end of March 2014. Also, the green shaded rectangle (presence of animals) is located at times with low FR_CH4. Ln 404-405: I disagree with the assumption that CH4 fluxes vary only to a small degree; this has not been shown. In Ln 390 the authors even argue that the hysteresis in Fig. 5 is due to changes in local emissions. I suggest citing literature describing the emissions from animals; what is expected from the process level, e.g. do ruminants C3