Review of ‘Role of thermodynamic and turbulence processes on the fog life cycle during SOFOG3D experiment’,

by Cheikh Dione, Martial Haeffelin, Frederick Burnet, et al.

Summary:-

This paper presents an analysis of data collected during the SOFOG3D campaign in south-west France during 2019-2020. Four cases of fog, and their evolution are discussed with an emphasis on their adiabaticity, and how a conceptual model performs in nowcasting the evolution of each case.

The paper is relevant and interesting, but requires some significant clarification to the data and arguments presented. Some of the analysis presented is difficult to follow, overly complex, and the cause and effect of various processes may be confused. My recommendation is to publish after major revision.

Main points:-

• Description of instrumentation.

Regarding the Doppler Lidar data I would like to see more explanation of how the TKE is retrieved, with an estimate of uncertainty given. I believe that the Lidar must scan over a region of sky to retrieve 3D winds, which raises the likely-hood that air samples in the separate beams are not coherent. Can any independent verification of the calculation be presented here?

Secondly, since the Lidar beam is highly attenuated by liquid water, how much of the fog layer is actually sampled? The authors state between 40 and 220 m, but I believe this is the range of the lidar and not how far the Lidar can typically see into the fog? Other similar Lidars typically can see into around 100m of fog.

Regarding the Microwave radiometer, an uncertainty is given for absolute humidity, but not the LWP, which is the quantity presented in the figures. Please provide an uncertainty for LWP.

Whilst the temperature error of the MWR is quoted as 0.5 degrees for the region of interest, it is clear that the profiles appear highly smoothed in the vertical (compared to what we expect to see from e.g. a tethered balloon profile). This might lead to erroneous conclusions regarding stability and phase of the fog. Were other sources of temperature profiles explored, such as radiosonde, mast or tethered balloon, before using the MWR data? It would be clearer if only the lowest 300m were plotted in the MWR temperature profile plots, and also if fog top were indicated on them at each time.

• Data analysis. Below are examples where I found the presentation of results requires clarification.

L319: - ‘thermal turbulence’: you mean, ‘thermally driven turbulence’?

L389::- It looks from the figure showing the MWR data that fog starts becoming adiabatic from 2400 or 0100 hours which is inconsistent with the statement here?

L393:- According to figure 6f sigma w^2 reaches 0.04 by 0300 hours, much more than the figure quoted.

L418-423:- I doubt the conclusion made here. The assertion is that a phase change in the fog caused a reduction in observed LWP, the evidence being ‘frost’ seen on the balloon cable. It is common to see ice and rime on such things when temperatures are below freezing in fog due to contact-freezing, but this is not evidence of ice or snow in the fog itself. Ice does not generally form in clouds until temperatures become much lower than seen here.

Section 3.3:- generally difficult to follow. Examples below.

L455:- Fig. 8a indicates a weak inversion at 2100 before the stratus lowers into the fog.

L458-459:- How does slowing down the cooling create a thin layer of temperature inversion?

L476:- Turbulence levels are very low for this case so how was the transition driven by turbulence?

L458:- What is ‘sustainable dissipation’?

L489. Why would a warming allow a deepening of the fog layer? I suggest the dissipation of this layer can be more simply put: Increasing wind aloft brought warm drier air over the top of the fog that then mixed into it, evaporating fog droplets, reducing RLWP to negative values and causing the fog to lift into low stratus.

L518:- ‘triggering of the ultra-low stratus being the fog’. Why not just say this was ‘stratus fog’?

L576:- Low stratus is not fog.

See attached