This is an excellent paper scientifically and is also very well written.  The authors use observations of the Schumann Resonance to modify the existing lightning parameterization in the GEOS-Chem model.  The existing Price and Rind cloud-top height prediction scheme for flash rate, as modified by Murray et al. based on satellite lightning observations, produced an increase in flash rates over the 2013 to 2021 period.  Correcting the parameterization using the Schumann Resonance yielded no significant trend, which is in line with observations.  The authors used the revised scheme in GEOS-Chem simulations for the period, examining the effects of the change in lighntning scheme on NOx, O3, and OH.    The updated scheme does better at producing interannual variability in these species.  The authors also examine the effects of a 10% decrease in lightining in 2020 on methane growth as a result of the decreased OH.  I have some suggestions for minor changes.  Once they are attended to, the paper should be ready for acceptance.

Details:

lines 121-125:  It would be worth noting that also the results of Allen et al. (2019, JGR) and Bucsela et al. (2019, JGR) from use of OMI NO2, also found little difference in LNOx production per flash between midlatitudes and tropics.

Figures 2, 4, 5, 6, and 7:  It is difficult to tell the gray and blue bars and lines apart.  A different color is needed for one of them.

Figure 6:  This figure needs better explanation.  From this figure I don't see how the non-lightning contribution is a difference between the observationa and the model lightning contribution.  For 2020, how do these bars imply a 54% contribution of lightning to the methane growth?  Why is there a methane growth rate bar for 2021, but not the lighting and non-lightning bars?

lines 271-273:  There is a significant amount of fire in South America during September to November.  I think you should choose a different set of months (maybe December to February).

Figure 7 - cation:  "green" should be "gray"

In this manuscript, Schumann resonance observations are used to create a scaling factor that further constrains the typical GEOS-Chem lightning parameterization based on cloud-top heights and the climatology of satellite observations of lightning. The trends and variability in global lightning activity, LNO_xemissions, O₃, and OH are then compared between GEOS-chem model runs with the Schumann resonance constraint, GEOS-chem model runs without the Schumann resonance constraint, satellite lightning observations, and ground-based lightning observations. Overall, the paper is well organized, and the incorporation of Schumann resonance data is a novel approach to modeling lightning and lightning chemistry. However, there are a few questions I would like to see addressed before I recommend publication.

Main comments

Line 91 states: “…and manually exclude the disturbed days (usually manmade activity) from the dataset.” Could you provide more information about the disturbed days, e.g. How do you know when a day is disturbed? What are the causes of disturbed days (i.e. what specific manmade activities lead to disturbances)?

Line 94: How does applying a third-order polynomial fit on the data remove the influence of solar cycle variations?

Figures 4, S1, S2: Could you explain a little more how the correlation coefficients in these figures are determined? For example, in Fig. 4a, how is the OP lightning contribution positively correlated with the non-lightning contribution if the OP lightning LTT is positive and the non-lightning LTT is negative?

Line 217: “The sharper decrease predicted by the SR simulation is in better agreement with the trends of tropospheric NO₂ columns from satellite observations during the period (Fig. 5a-b).” This interpretation does not seem consistent with the Figure 5 for the following reasons:

• Looking at Figure 5, the NO₂ anomalies determined from the SR method appears much more similar to those from the OP method than to the OMI observations. For example, both the OP and SR results underestimate the 2015-2016 NO₂ anomalies and overestimate the anomalies in 2018-2019 by about the same amounts relative to the OMI data. Thus the SR simulation does not substantially improved the agreement with the OMI data compared to the OP simulation. Are the differences between the OP and SR trends (-5.4% [SR] vs. -3.8% [OP] for global and -10% [SR] vs. -8.3% [OP] for the Northern Hemisphere) significant?
• It is true the magnitude of the SR trend (-7.0) is closer to the OMI trend (-7.1) than to the OP trend (-5.4) in Figure 5a, but the OMI trend is not significant, while the OP and SR trends are both significant. Again, this result suggests that the SR results are not really capturing the OMI trend any better than the OP results.

Perhaps the conclusion is that some other factor besides lightning is the reason for the differences between the model results and satellite observations for the NO₂ anomalies?

Fig. 7a: What is causing the positive change in NO₂ anomaly for the OP method? According to Fig. S3a, LNO_x determined from the OP method also decreases from 2019 to 2020, just not as much as the SR method LNO_x. So it is a little surprising to see a positive NO₂ anomaly from the OP method in Figure 7a.

Throughout the paper the trends are often described as decadal or the changes put in terms of % per decade, but the study range only covers 8 years of data, so describing the trends as decadal (occurring over 10 years) seems not quite right. Perhaps “overall” trend or similar wording would be more appropriate?

Technical Comments

Line 54: “…lightning observations from satellite-based the Lightning Imaging Sensors (LIS) and Optical Transient Detector (OTD)…”. The words “the” and “satellite-based” should be switched: “…lightning observations from the satellite-based Lightning Imaging Sensors (LIS) and Optical Transient Detector (OTD)…”.

Line 87: “Geoagnetic” should be “Geomagnetic” (“m” is missing).

Both Line 192 and Line 257 use similar phrasing that is difficult to follow: “occurs in a decrease” and “occurs in a huge decline”. Simplifying the wording here would make these sentences flow better, e.g. “…global lightning activity decreased by ~10% from 2019 to 2020 (Fig. 2a).” and “…global lightning declined by ~10% from 2019 to 2020.”

Supplement: There is a typo in first sentence of Text S1: “mothed”- maybe should be method?