The manuscript delivers a comprehensive exploration into the creation of a global dataset that delineates the vertical structure of clouds, harnessing the combined observational capabilities of CloudSat and CALIPSO. The attempt to integrate data from these two satellites is both ambitious and timely, providing a unified perspective on cloud distribution, altitude, and overall cloud architecture. The result is a dataset that could become an invaluable tool for climatologists, meteorologists, and researchers working in atmospheric sciences. While the paper offers an impressive scope, a few suggestions for enhancement are presented below:

Specific Comments:

1. In Figures 6 and 9, the portrayal of global cloud patterns and anomalies is instructive. However, it would enhance the paper's depth to delve into more elaborations about regional variations, the role of seasons, and potential correlations with established climatic phenomena. In addition, the authors may discuss whether the datasets can potentially illuminate some new insights into cloud characteristics.
2. It's laudable that the authors have illuminated potential discrepancies and constraints within the dataset in the error analysis section. Nonetheless, a more expansive discussion on the inherent uncertainties, especially when navigating regions marked by different topographies or areas known for complex cloud systems, would be invaluable.
3. For clarity and utility, it might be beneficial to delineate exact numbers related to the uncertainties inherent in datasets across different global regions. Such specificity would aid users in gauging the reliability of the data in diverse contexts.
4. One notable observation is the narrow focus on dataset evaluation at Utqiagvik, Alaska, despite its global ambit. Given the universal scope of the dataset, is there a reason for concentrating evaluation efforts in this isolated region? With resources like the ARM program, which boasts a spectrum of global observatories, a broader evaluation could reinforce the dataset's credibility across multiple geographies.

This manuscript describes a new data product which reports monthly, globally gridded 3-D cloud occurrence based on co-located observations from the CALIPSO and CloudSat satellites.  The manuscript presents a comprehensive description of the product and a detailed comparison of the product with results from CALILPSO-only and CloudSat-only which helps in interpretation of the joint results.  The manuscript does a good job of describing the product and its strengths and limitations, but the manuscript would benefit from clarification of a few points.

The discussion in lines 31-49 of detection sensitivities of radar and lidar is a little misleading in casting the comparison in terms of layer optical depth.  The differences in detection sensitivity between the two instruments are primarily driven by particle size and not as much by concentration.  Liquid clouds such as stratocumulus and altocumulus (composed of relatively small droplets) are often undetected by CloudSat even though they can be opaque to CALIOP.  Small, cold ice particles detected by CALIOP are often missed by CloudSat, or CloudSat may only detect the drizzle component of these clouds.  But CloudSat can detect low concentrations of very large ice particles (perhaps precipitating) which are not detectable by CALIOP due to the low concentrations.  This is nicely characterized in Liu et al. (2016) and Liu et al. (2018).  Another thing which is a little misleading: CloudSat surface clutter is due more to sidelobes of the radar beam than to the pulse length.  CALIOP does  not suffer from the same phenomenon not only because the laser has a much shorter pulse but it also has no sidelobes, due to the wavelength being three orders of magnitude shorter.

The sentence in lines 104-106 is a little confusing where it says that 2B-GEOPROF-LIDAR uses a “fixed maximum along-track averaging length of 5 km for cloud detection”.  This sounds like maybe 2B-GEOPROF-LIDAR is performing its own cloud detections using 5-km averaging.  My understanding is that 2B-GEOPROF-LIDAR is using only the layers from the CALIPSO Layer Product which have been detected in 5-km averages and ignoring the layers in that product which are detected after additional averaging, but I could be wrong.  This point should be clarified.

Line 130: The authors clearly explain here that ‘lidar attenuation’ is determined by comparing the CALIOP and CloudSat cloud masks.  However, this may result in errors when CloudSat fails to detect a cloud which is opaque to CALIOP.  This is likely responsible for some of the differences shown later in the paper.  In a future version of this product, the authors might consider using the presence or absence of lidar surface detection to identify lidar full attenuation, which is the approach used for CALIOP Level 2 products.

Line 175: In partitioning cloud cover into High/Middle/Low levels the authors use a fixed altitude to approximate the ISCCP pressure levels, which introduces a source of uncertainty in comparisons between this product and results from passive sensors.  It is straightforward to compute monthly zonal-mean altitudes corresponding to the ISCCP pressure levels, which mostly eliminates these uncertainties.  This is the approach used for CALIPSO-ST and MISR in the GEWEX Cloud Assessment and authors should consider this approach for a future version of the data product.

Line 200: “latitudes of the first and last DO-Op profiles” is mentioned here and in several places later in the manuscript.  It wasn’t clear to me what is meant by ‘first’ and ‘last’, please explain.  

Line 306:  The statement that CAL-COS ‘excludes marine low clouds detected only on 5 km averaging lengths’ is not very clear.  What is done for marine low clouds in CAL-COS is that only single-shot detections of shallow liquid clouds are used and all the results based on horizontal averaging are ignored.  We have realized single-shot detection is sensitive enough to detect virtually all of the shallow liquid clouds and that using layers detected in 5-km averages tends to overestimate shallow cloud cover.

Minor comments:

Line 38:  For me “level of attenuation” is ambiguous, as ‘level’ could refer to signal magnitude.  I think ‘altitude of attenuation’ is what is meant.

Line 57: I did not find a Figure 4.2c.  Should this be Figure 2c?

References:

Liu, D., Q. Liu, L. Qi and Y. Fu, 2016: “Oceanic single-layer warm clouds missed by the Cloud Profiling Radar as inferred from MODIS and CALIOP measurements”, J. Geophys. Res. Atmos., 121, 12,947–12,965, https://doi.org/10.1002/2016JD025485.

Liu, D., Q. Liu, G. Liu, J. Wei, S. Deng and Y. Fu, 2018: “Multiple factors explaining the deficiency of Cloud Profiling Radar on detecting oceanic warm clouds”, J. Geophys. Res. Atmos., 123, 8135–8158, https://doi.org/10.1029/2017JD028053