The article investigates model configuration strategies to improve the accuracy of atmospheric greenhouse gas (GHG) transport simulations using the Weather Research and Forecasting (WRF) model coupled with the GHG module (WRF-GHG). The study focuses on minimizing transport errors by coupling WRF-GHG with ERA5 meteorological reanalysis data through two main strategies: daily model restarts and continuous grid nudging. Six experiments were constructed by applying different configurations of these strategies: NN_NR, NN_DR, GN_NR, GN_DR, GN_3km_NR, and GN_3km_DR through a two-month-long simulation over the European domain. The authors compared the model output with both meteorological and CO2/CH4 measurements and concluded that (1) both daily restarts and grid nudging improved meteorological accuracy and GHG transport, with a small advantage when both methods were combined; (2) notable differences in soil moisture were observed, which accumulated over the simulation period when not using frequent restarts. This drift in soil moisture affected the simulated planetary boundary layer height (PBLH) but did not significantly impact GHG performance; (3) Daily restarts or nudging minimized soil moisture drift, enhancing the simulation of surface temperature and humidity, and improving PBLH representation.

This work provides a strong, logical recommendation for the WRF-GHG setup and well-documents the results. I have been using this strategy for my simulations but haven’t done or seen any work illustrating the rationale behind it. The study also provides valuable insights into optimizing long-term tracer simulations with WRF-GHG. By recommending a combination of daily restarts and grid nudging, the study offers a practical solution to enhance the accuracy of atmospheric GHG transport models, contributing to better quantification of inversion. This article has done thorough work in terms of model evaluation. The method is sound, and the results and conclusions are solid. I have no concerns regarding language or grammar. I would recommend this article after the authors address my specific comments below.

Specific Comments:

• In the introduction, I noticed that no literature was cited beyond 2018. This raises a question about whether there has been no relevant work in this area over the past six years, which I found quite surprising. I recommend considering the inclusion of the following recent citations to provide a more comprehensive overview of the current state of research in this field:

Feng, Sha, Thomas Lauvaux, Kenneth J. Davis, Klaus Keller, Yu Zhou, Christopher Williams, Andrew E. Schuh, Junjie Liu, and Ian Baker. “Seasonal Characteristics of Model Uncertainties From Biogenic Fluxes, Transport, and Large-Scale Boundary Inflow in Atmospheric CO2 Simulations Over North America.” Journal of Geophysical Research: Atmospheres 124, no. 24 (2019): 14325–46. https://doi.org/10.1029/2019JD031165.

Feng, Sha, Thomas Lauvaux, Klaus Keller, Kenneth J. Davis, Peter Rayner, Tomohiro Oda, and Kevin R. Gurney. “A Road Map for Improving the Treatment of Uncertainties in High-Resolution Regional Carbon Flux Inverse Estimates.” Geophysical Research Letters 46, no. 22 (2019): 13461–69. https://doi.org/10.1029/2019GL082987.

Gerken, Tobias, Sha Feng, Klaus Keller, Thomas Lauvaux, Joshua P. DiGangi, Yonghoon Choi, Bianca Baier, and Kenneth J. Davis. “Examining CO2 Model Observation Residuals Using ACT-America Data.” Journal of Geophysical Research: Atmospheres 126, no. 18 (2021): e2020JD034481. https://doi.org/10.1029/2020JD034481.

• I find the justification for evaluating modeled GHG against observations to be lacking. Feng et al. (2019a; 2019b) demonstrated that flux uncertainty predominantly influences model CO2 uncertainty. This likely explains why the CO2/CH4 simulations from different experiments do not exhibit as much distinction as the meteorological variables. The authors should address this issue and provide a stronger rationale for their approach.

The authors may include these two references when they explain the causes of the similar performances in terms of GHG simulations in Line 315.

• The content and significance of Figure 8 are unclear. Regardless, I observe a consistent trend in model errors across the different flights. It appears that the reanalysis data, which provide the initial and boundary conditions for WRF-GHG, dominate the model errors. The authors should clarify the information presented in Figure 8 and discuss the impact of reanalysis data on the model’s performance.
• The p-value presented in Figure 10, such as 1.24e-09, appears questionable. Could the authors clarify the number of samples used for the statistics in Figure 10? Are those r values meaningful?

In this paper, the WRF model is used to simulate the transport of CO2 and CH4 for a 2-month period in 2018 at 5-km scale over Europe, with particular focus over a coal mining region of southern Poland. The authors compare the use of grid nudging and model reinitialization to improve the representation of the near-surface temperature, humidity, winds, and PBL height, and assess the implications for simulating GHGs. While somewhat similar studies have been conducted previously comparing nudging and reinitialization for simulating regional meteorology and climate, none to my knowledge have been at such high spatial resolution, nor have they focused on pollutant transport. Additionally, the linkage between soil moisture drift and PBLH errors is novel. The manuscript is well constructed and clearly written.

I recommend the paper be accepted for publication after the minor issue noted below is clarified or corrected.

Lines 271-274 say that flights suitable for model-data comparisons are denoted with asterisks in Fig. 8. However, the caption to Fig. 8 says that asterisks indicate "flights that crossed so close to nearby point sources that we cannot represent them well." It appears that the authors at some point changed whether asterisks denoted "good" vs "bad" flights for comparison against the model.