Some general comments

The paper demonstrates that including the response of large-scale subsidence to changes in the mean vertical buoyancy profiles is crucial for assessing the strength of the cloud response to salt spraying in the stratocumulus to cumulus transition (SCT) regime. It may be helpful to give more attention to explaining the choices made in setting up the vertical profiles of aerosol concentrations (Na) according to Table 1, particularly since Na is relatively low in the boundary layer and higher in the free troposphere, which may be typical of the location studied in the paper. Additionally, as the study on the interaction between boundary layer turbulence and large-scale circulation is performed using a high-resolution large-eddy simulation model on a small domain, could the authors briefly discuss the pros and cons of this approach compared to using a global model, such as the one applied by Wan et al. (Nature, 2024)?

(Minor) remarks

line 117: Include a reference to the Morrison microphysics scheme?

line 135: The vertical profiles of the buoyancy do generally depend on time and space, and according to the Lagrangian setup of the simulations, the position of the LES domain changes with time. However, it is not immediately clear whether the 'diagnosed buoyancy profiles' are taken constant during the simulation, or that they depend on the position (and local time) of the LES domain? The results suggest the latter, in particular the caption of Fig. A1 states that "The red line represent(s) the ERA5 climatology at the location of climatological trajectory on Day 2" (note that Fig A1b does not show a red line). All the relevant details on the WTG application are present, but I find them a bit scattered around in the manuscript. As a side note, the main text mentions "domain-mean anomalies of virtual temperature and diabatic heating with respect to diagnosed buoyancy profiles", while Appendix A uses the virtual potential temperature. Perhaps include their definitions?

Line 247, the buoyancy flux used is different from its definition <w'b'> = g/ <w'theta_v'> (with <> indicating a slab mean value). The statement that "changes in SHF play a leading role in changes in the surface buoyancy flux in these simulations" seems a bit too simplistic. The SHF and LHF values in the Sandu cases are on the order of 10 and 100-200 W/m2, respectively (see Sandu and Stevens 2011). Because the LHF can be an order of magnitude larger than the SHF, the term 0.07 x LHF can become as large as the SHF.

2.3 Data. N_a is not defined (first occurence in the text, line 156).  Could the initial values for N_a in the MBL and FT presented in Table 1 be motivated? In particular the lower values of N_a in the MBL compared to the ones in the free troposphere?

Line 157, Spin up procedure. An 18 hour spin-up period is applied. Can some more details be given about how this is done? For example, are the large-scale conditions set constant or not during spin up, is nudging applied?

Line 158, can the choice of the aerosol injection rate be motivated? Does it involve evaporative cooling of the sprayed water?

Line 163, rotation of the domain. Due to surface friction and the resulting momentum fluxes the wind in the boundary layer will turn with respect to the geostrophic wind direction. However, it is stated that (with a rotated domain) 'the x component of background wind velocity is approximately zero',  which suggests that the ageostrophic component of the wind vector is close to zero. Please clarify.

Line 177, Figs 2 and 3, f_c is defined as a cloud fraction but also as a cloud cover. The cloud fraction is often defined as the ratio of cloudy area to the total area on a horizontal plane as is applied in Fig. 3, but Fig. 2 shows the cloud cover (ratio of cloudy air columns to all vertical columns).

Line 188, 'intensified subsidence .. delivers more CCN into the MBL' . Aren't CCN entrained into the MBL, subsidence is just pushing down the boundary layer?

Line 279, 'turbulence dissipation by the decreased surface buoyancy flux'. The sign of the buoyancy flux in the subcloud layer matters in this respect, at heights where it is negative it will tend to diminish turbulent kinetic energy.

Line 286, Can you explain what is meant with 'the aerosol-cloud interaction is not yet saturated'?

Appendix A. The WTG application is strongly based on ERA5 fields. Given the typical biases in weather and climate models of various quantities in the SCT regime, it is maybe worthwhile to briefly discuss the accuracy of the vertical profiles of the thermodynamic variables in the SCT regime in ERA5?

Fig. 2: refer to 'large-scale' vertical velocity in Fig. 2b. Explain the meaning of the grey bands?

Typos

line 105: "Arkawa"