Summary of manuscript

The authors present observations and their basic statistical evaluation of meteorology, boundary layer properties, and cloud properties collected over the period 2013-2018 at the Atmospheric Radiation Measurement (ARM) Facility's Eastern North Atlantic (ENA) site on Graciosa Island. The data are divided into two categories based on the state of cloud mesoscale organization: closed-cell stratocumulus and open-cell stratocumulus. Properties of composites of the data (mean, variability, diurnal cycle, probability distribution functions) in each category are presented and discussed.

Summary of review

This manuscript gives an excellent, quantitative, systematic, and detailed view of the closed- and open-cell stratocumulus state and of their meteorological context at Graciosa Island. The text is written clearly, logically, and consisely, and the data and their analysis are presented in a way that will make them very useful for others. This work hence advances the science, even though no specific new scientific insights are presented.

There are a few items, listed below, in the text that require attention before publication. Thes items are significant in substance and should not be glossed over, but can be addressed with a focused yet limited effort. Hence a minor revision is recommended.

Specific comments

Line 42-42, and 316-317: "Despite these important impacts, the atmospheric processes responsible for the formation and maintenance of these organizational states remains poorly understood."

A host of literature exists on the processes that are involved in driving and maintaining closed- and open-cell stratocumulus cloud decks, and the processes that convert one into the other. Hence writing that "processes responsible for the formation and maintenance of these organizational states remains poorly understood" is incorrect. Here are a few references. Please add any other works based on a literature search and change the text to reflect the existing understanding of the subject.

Shao and Randall, 1996, Closed Mesoscale Cellular Convection Driven by Cloud-Top Radiative Cooling

Stevens et al., 2005, Pockets of open cells and drizzle in marine stratocumulus

Comstock et al., 2005, Mesoscale variability and drizzle in Southeast Pacific stratocumulus

Wood et al., 2008, Open cellular structure in marine stratocumulus sheets

Bretherton et al., 2010, Southeast Pacific stratocumulus clouds, precipitation and boundary layer structure sampled along 20° S during VOCALS-REx

Wood et al., 2011,  An aircraft case study of the spatial transition from closed to open mesoscale cellular convection over the Southeast Pacific

Xue et al., 2008, Aerosol effects on clouds, precipitation, and the organization of shallow cumulus convection

Savic-Jovcic and Stevens, 2008, The structure and mesoscale organization of precipitating stratocumulus

Wang et al., 2009, Modeling mesoscale cellular structures and drizzle in marine stratocumulus. Part II: The microphysics and dynamics of the boundary region between open and closed cells

Feingold et al., 2010,  Precipitation-generated oscillations in open cellular cloud fields

Kazil et al., 2014, On the interaction between marine boundary layer cellular cloudiness and surface heat fluxes

Yamaguchi et al., 2015, On the relationship between open cellular convective cloud patterns and the spatial distribution of precipitation

Feingold et al., 2015,  On the reversibility of transitions between closed and open cellular convection

Line 54-56: " However, Wood et al. (2011) and Terai et al. (2013) found that drizzle rates are not significantly different between open and closed cells, concluding that drizzle, and its associated thermodynamic feedbacks, are not the only factor causing the transition between mesoscale organizations."

There is more to be written on this subject - see, e.g., Yamaguchi et al., 2015, "On the relationship between open cellular convective cloud patterns and the spatial distribution of precipitation", and Feingold et al., 2015,  "On the reversibility of transitions between closed and open cellular convection".

Line 63-65: "Hence, most of the open-cellular stratocumulus that are observed over the ENA are fundamentally different than those observed in the other parts of the subtropical oceans under quiescent large-scale forcing conditions"

Cold air outbreaks do conceptually differ from the air masses in the eastern subtropical oceans with stratocumulus cloud decks, but do the open cells really differ in these two situations, and how? In other words, it will not do to just state that "open-cellular stratocumulus that are observed over the ENA are fundamentally different than those observed in the other parts of the subtropical oceans under quiescent large-scale forcing conditions" without supporting evidence, literature references, and specifics. Please provode these or remove this passage if it cannot be supported.

Figure 8e: A logarithmic y-axis might perhaps help visualize the data better(?)

This study investigates the difference between closed- and open-cellular low clouds using long-term ground observations from the ARM ENA site on an island of the Azores.  The biggest merit of the study is its usage of long-term data (~ more than 5 years) whereas previous studies of the topic are mostly based on short-duration field campaigns. The manuscript is very well written. The technique is overall sound. For these reasons, this paper has great potential to be published in ACP. 

However, there is a lack of new insights. Most findings are consistent with prior knowledge. The presentation of the results is too descriptive and lacks discussions about the underlying physics. Also, some conclusions are questionable due to insufficient reasoning. These issues must be addressed. See below for detailed comments.

Major comments:

1.    It is not straightforward from Figure 4 to conclude that "cold air advection is stronger for open-cellular cases". It seems the majority of trajectories in open-cell cases align with the SST contours, indicating a weak temperature gradient along the trajectory.  Quantification is needed, which is absent in the analysis. 

2.  The analyses of the composite sounding have a couple of problems. First, the height should be scaled (e.g. by the main inversion height) to avoid artificial cancellation. Second, similarly, scaling may be needed for the temperature and humidity because samples are from different seasons (e.g. scaled by the seasonal mean). Given the limited number of cases, the composite absolute values of both variables may be biased toward a specific season with the largest samples.        

3. The discussion about the time evolution of the thermodynamic structure (Fig. 6) lacks physical insights. How do the diurnal cycle and advection influence evolution? 

4. Line 198. Could you discuss why closed-cell cases are more coupled? Is it due to large-scale factors such as temperature advection or the boundary-layer scale factor such as cloud-top radiative cooling or precipitation?  

5. Lines 222-223. Again, why open-cellular cases are associated with stronger subsidence ) needs more discussion. From the perspective of PBL mass conservation: dzi/dt = ent - wsub, in which zi is the inversion height, ent it entrainment rate, and wsub is subsidence rate. Larger subsidence leads to PBL swallowing whereas open-cellular cases have deeper PBL. Therefore, it must be the stronger entrainment that is responsible for the deeper PBL.  

6. Line 230. Sill, why open-cellular cases have slightly higher PWV should be discussed. The key idea of Zhou and Bretherton's series of studies on cellular convection is that PWV perturbation is highly correlated with LWP perturbation. For case-averaged PWV, the story is completely different. The latter is a synoptic-scale problem whereas the former is a mesoscale problem.      

Minor comments:

Figure 1: the resolution is a little bit too low. 

Figure 9: use local time or mark the period of day/night times.