see attached PDF file

This work of Rosenberg et al. discusses a new approach to analyze higher-order moments of turbulent fluctuations in the convective boundary layer (CBL). Instead of normalizing the height range against the mean depth of the CBL in the analysis period (typically 30 minutes to 2 hours), the authors suggest a height grid normalized with a temporally smoothed depth of the CBL.

The authors base their analyses on large eddy simulation (LES) data of one day and discuss variance, skewness, and kurtosis of the turbulent fluctuations of water vapor mixing ratio as well as of vertical wind, and of the covariance of these two quantities, the latent heat flux (water vapor flux). Furthermore, they compare these simulated measurements of a vertical pointing lidar (one column measurements) with the data averaged over the whole domain of their model.

The manuscript is well written. I recommend accepting the manuscript after minor revision.

Specific comments:

Page 4, line 70ff: I think it would help if the authors added formulas to explain how they calculated the parameters for the three methods (more than just text). In addition, I suggest adding example plots of the three methods (or at least of the two lidar simulations) for a 1-hour example.

Introduction: Please add references to the statements in the first paragraph.

Figure 1: Please use the same nomenclature for the different parameters like in the text (line 74). ‘Blue’ should be ‘cyan’, ‘green’ is ‘black’. Mark local noon, sunrise, and sunset.

All figures: Please add labels (a), (b), (c).

Figure 2ff: Please define the white areas. I would prefer units at all color bars (instead of explaining the units of the color bars in the figure captions) and labels with the full parameter information (not just “Variance”, “Flux” etc.).

Table 1, 2, 3: You refer to RSME values but do not show these. I suggest that you also show the RSME data in the tables.

I think it would be better to use the same color scales for all plots of a figure, not different ones, so that it is easier to compare the results (Figs. 2, 6, 7). Same for the y scales in Figs. 8, 9, 10.

I would also prefer time scales with ticks at 6, 12, 18 h etc.(not at 5, 10, 15, 20 h). In addition, small ticks at each hour would be helpful.

Instead of “error”, I would prefer the term “uncertainty”.

Instead of “q flux” (Fig. 6) or “WVMR flux” (Fig. 10), I would prefer “latent heat flux”.

All units should be in normal font (not italic).