Precipitation over the U.S. Coastal Land/Water Using Gauge-Corrected Multi-Radar/Multi-Sensor System and Three Satellite Products
Round 1
Reviewer 1 Report
I have gone through the manuscript, and I find it reads well and easy to follow. The presentation is in good shape, and this study provides a comprehensive analysis of the land/sea contrast of coastal zone precipitation from a satellite perspective. Using the ground observation data (MRMS) they detect a significant difference in land and ocean precipitation. This difference is then verified in three major satellite precipitation products (IMERG, PERSIANN, and CMORPH) in a three-year window. They found that IMERG works best over land while CMORPH works better over ocean. In addition, PERSIANN has an overall best presentation of such land/sea contrast. With the impact of the uncertainty in MRMS examined, these conclusions are pretty robust and straightforward recommendations on the future use of satellite produce in studying coastal zone precipitation. I would recommend publishing after minor revision.
Specific comments:
1. Maybe we want to discuss the impact of the thee-year time window. As far as I am aware, all these dataset provides more data beyond 3 years. So some justification for only focusing on such a short period might be necessary. I guess there are significant extremes in this window, which can partially justify this choice. But I think a further discussion can be very helpful.
2. The discussion on the cause of diverse (and interesting) performance of three products can be highlighted. While they are mentioned in conclusion (line 328-333), the details are buried in line 261-276, and I think they are worthy of further highlighting. These are pretty useful suggestions for the data product teams. Meanwhile, besides the bias correction algorithm, are these differences dependent on the sensor types and (possibly different) precipitation types over land versus ocean?
Author Response
Thank you for your comment! Please see attached file.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Overview:
This manuscript presents an important evaluation work about satellite precipitation products. I would consider the topic of this work fits the scope of Remote Sensing very well and has a high degree of scientific merit. Precipitation products from satellite have a unique advantage—the global coverage, however, they are underutilized because of a lack of systematic evaluations. I am glad to see this work integrate to fill the research gap and would let more scientists have a more clear picture of the biases and uncertainties in the products. The manuscript is overall well written and the figures are delivering the message. While I do have a couple of comments regarding the overlook of the role of atmospheric rivers on the West Coast precipitation and the lack of discussion that links to findings in the literature, I would recommend a minor revision to allow this important work to be published sooner.
L11: Change “rainfall” to “precipitation”
L15: One might want to spell out these product names.
L17: How about the springtime?
L45–52: I would suggest adding a few lines to summarize the overall biases of each product type.
Figure 1: Caption: one might not want to say 1a is the climatology as this is only from 2018–2020.
R99p should be defined somewhere in the method section or in the text for discussing Figure 1b.
L173: Atmospheric rivers are known to be one of the contributing factors to providing sufficient water vapor to enhance the precipitation and latent heat release, which is critical to the heavy precipitation over the topography. About 18% of ARs are independent of cyclones and drive heavy precipitation along the US West Coast (Zhang et al. 2019). One should not ignore landfalling ARs when discussing the systems for delivering precipitation along the US West Coast. See the following paper about the dynamical relation between ARs and cyclones:
Zhang, Z., Ralph, F. M., & Zheng, M. (2019). The relationship between extratropical cyclone strength and atmospheric river intensity and position. Geophysical Research Letters, 46(3), 1814-1823.
Figure 1c: One might want to clarify what cross-section this is such as from ?? to ???
Section 3.2: There is a lack of discussion to compare the results in this study and those in the literature. For example, are the IMERG-F biases over land, which seem to be the smallest among all products, consistent with previous studies or any indications from previous studies? What are the differences, if any, from previous studies?
Section 4: Stage-IV precipitation products have been widely used for evaluation over CONUS. The whole manuscript did not mention it at all. One at least wants to briefly mention this dataset. See the following reference for more detail:
Du, J. 2011. NCEP/EMC 4KM Gridded Data (GRIB) Stage IV Data. Version 1.0. UCAR/NCAR - Earth Observing Laboratory. https://doi.org/10.5065/D6PG1QDD. Accessed 30 Aug 2022.
Author Response
Thank you for your comment! Please see attached file.
Author Response File: Author Response.pdf
Reviewer 3 Report
Please see attached pdf.
Comments for author File: Comments.pdf
Author Response
Thank you for your comment! Please see attached file.
Author Response File: Author Response.pdf
Reviewer 4 Report
This review is about the article "Precipitation over the US Coastal Land/Water Using Gauge-Corrected Multi-Radar Multi-Sensor System and Three Satellite Products" by Yike Xu et al.
I consider the article very well written. The language is easy to understand, the analysis is statistically sound, and the scientific interpretation - when comparing these different products - is very impartial and objective.
I will propose to accept the article in its present form. However, if the editor or my fellow reviewer(s) don't agree, I would have a few suggestions that might help to understand the analysis better. I want to point out that these are suggestions only, and I leave it to the authors to add them or not.
Data analysis and Methods:
I had to jump between this chapter and the "Results and Discussion" to remember the main differences between the products. Especially to remember sources (IR or MW) or the fact, which one applies gauge correction or not. It would be beneficial if the authors could add a little table at the end of the chapter that summarizes the main difference (resolution, main satellites, gauge correction, ...)
Results and Discussion:
Figure 1c misses an x-Axis label. It can be referred from the text and other Figures afterward. But as it is the first of its kind, it is slightly confusing. I would suggest adding a label.
The other thing that bugs me a little bit: N, the number of samples per distance bin, which seems to be decreasing with increasing distance from the coastline for MRMS, but seems to increase for some of the products. Figure 3 gives a good hint about that. However, the color scheme makes it hard to verify: for MRMS, white seems to mean "no values." However, for the products, white appears to represent "no values" and "100 values". It's, therefore, hard for readers to estimate the "total N" per distance bin. I would propose changing the product color scale and perhaps adding a plot with "total N."
Also, I would propose a "log/log"-plot for the correlations plots in Figure 4. Most points are in the lower range, making the correlations easier to see.
Conclusions:
The analysis is very focused on numbers, but the conclusion seems to miss these. A Lot of people jump directly from the abstract to the conclusion, so I would propose adding a few numbers to that part to verify your results. That's not mandatory, but it helps people to understand your conclusions better.
Again, I consider none of these suggestions mandatory; I think the article is good enough as it is now. They are merely ideas to be kept in mind if the editor or the authors consider a revised version.
My Regards.
Author Response
Thank you for your comment! Please see attached file.
Author Response File: Author Response.pdf
