Summary

The paper addresses the critical importance of accurately estimating global precipitation, especially at the sub-daily scale, where uncertainties are typically higher than daily, monthly, or annual estimates. The study has been conducted on satellite-based and reanalysis products on a global scale. The study specifically investigates the diurnal cycle of precipitation using a comprehensive analysis of five global precipitation products that provide at least hourly resolution data: The analysis focuses on three key parameters of the diurnal variability of precipitation: precipitation amount, frequency, and intensity. The study highlights the need for integrating diverse datasets to minimize uncertainties and ensure a more accurate analysis of global precipitation patterns. Relying on a single precipitation product can lead to misrepresentations due to the inherent discrepancies in the diurnal cycle estimates.

General Comments

The study addresses a quite interesting topic. The paper clearly defines its objective of evaluating sub-daily precipitation estimates and analyzing uncertainties. It emphasizes the importance of using diverse datasets for precipitation analysis, stressing caution in relying solely on individual products. The manuscript is well organized and neatly written with the appropriate scientific content. However, I have some suggestions and questions as follows:

Major Comments:

• If the primary goal is a relative comparison among the products (e.g., to understand differences in product outputs without focusing on absolute accuracy), then comparing the products directly to each other is acceptable. However, for a comprehensive evaluation of accuracy and applicability, it is ideal to use a reliable benchmark dataset. This provides valuable insights into the behavior of each product with respect to the intensity and frequency, allowing for a more robust assessment.
• Why did you not use the latest version of the IMERG product (v07) in this study? Using the most recent version is crucial, as updates in IMERG v07 may include improvements in algorithm accuracy, enhanced calibration, and adjustments based on newer gauge corrections. These enhancements can significantly impact the reliability and precision of precipitation estimates, which is particularly important for a robust inter-comparison of precipitation products. I strongly encourage you to consider using IMERG v07 that has been released in 2023, in your analysis to ensure that your findings are based on the most accurate and current data available.

Minor Comments:

• For PERSIANN, and CMORPH the link where the data have been downloaded are mentioned, please do the same for IMERG and GSMaP
• Line 136-139: please add a reference
• Line 170-172: This statement is not fully clear to me. Does it mean that in this case, adding more than 4 clusters improved WSS by less than 7%? And based on that the additional clusters did not significantly enhance clustering quality. Selecting k=4, avoids overfitting while preserving meaningful pattern distinctions in the diurnal precipitation data. Is it correct? If yes, I suggest rephrasing your statement to be more clear.
• Line 216-217: Please mention which index you are referring to, here you refer to intensity, am I right? Do you have an idea of what might be the possible causes for the sharp decline of GSMaP over south hemisphere?
• Line 284-288: Is the only possible reason for differences between IMERG and GSMaP the use of different gauge corrections? What about the variations in their underlying algorithms? In short, while both IMERG and GSMaP utilize a similar constellation of satellite estimates, they differ in algorithms, processing techniques, and data sources, all of which could contribute to the observed discrepancies in precipitation estimates.
• I’m also interested in learning more about any overlap between the GPCC and CPC networks, as well as an approximate count of gauges used by each. Additionally, since GPCC and CPC likely apply different quality control, calibration, and interpolation methods, this could further contribute to differences in their final precipitation products, even when some gauges are shared.
• Line 303-306: It might be interesting for some readers to have more information on some of the possible reasons for these phenomena! For example, the interactions between Sea Surface Temperature and Boundary Layer: Over oceans, cooling at night allows for more stable conditions near the surface, which can promote the formation of precipitation in the early morning hours. The surface cooling also reduces convective inhibition, allowing for nighttime or early morning convection. You can also mention other possible reasons such as “latitudinal differences in sunlight and heat retention” etc.

            The same comment applies to the “coastal regions” and “over land” parts.

• Figures 6, 7, and 8: These are very good figures; they're highly informative and well-designed."
• Page 15, Figure 7: Do you know why GSMaP displays these north-south striping patterns for the peak precipitation frequency hour?"

Thank you for this interesting paper! Please find in the document attached my review. 

Best regards, 

Martina Kauzlaric