Evaluation of Warm-Season Rainfall Diurnal Variation over the Qilian Mountains in Northwest China in ERA5 Reanalysis

Round 1

Reviewer 1 Report

This study provides a detailed evaluation of the European Centre for Medium-Range Numerical Weather Forecasts reanalysis version 5 (ERA5) in the precipitation amount, frequency, intensity, and diurnal variation over the Qilian Mountains in Northwest China during May-October. The deviation in ERA5 comes from the complex topographic effects and related bias in surface specific humidity.

I think this manuscript could be further improved after some minor revisions.

Major comments:

1. In Data and Methods, the observation dataset covers the periods of May to October from 2012 to 2019, but why the author chooses May-October as the target period? How about the seasonal (monthly) evolution of precipitation over Qilian Mountains?

2. Figure 3. How to calculate the Alpha and Beta in Figure3? What is the physical meaning of Alpha and Beta?

3. In definition of the duration of precipitation events, why the duration time of all events are less than 24 hours? Why the spatial distribution of Qilian Mountains differs from other sections?

4. There are many grammar flaws and incorrect phrases, and they need to be corrected.

Minor comments:

1. Page 3 Line 96, “a horizontal resolution of 25 km” is “a horizontal resolution of 25 km×25km”.

2. Nearly all the paper. When describing the detail of a specific figure, Figure 1 for example, the marker should be (Figure 1), not just a “Figure”.

3. Page 4 Line 133, the RMSE should be replaced by NRMSE.

4. Page 7 Figure 4, the distribution of weak and heavy precipitation in ERA5 and observation should be shown. Same as Figure 9 and Figure 12.

5. Page 9 Line 263, (mm y^-1) is (mm day^-1).

6. Page 10 Line 267, “weak and heavy” is “(a) weak and (b) heavy”.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The need for reliable data for planning climate adaption measures is obvious while data scarcity in an issue in many regions. Therefore, the use of reanalysis data is more and more common and important – but: validation is required. Therefore, the submitted paper is relevant and appropriate for the journal.

The paper is generally well written and scientifically sound, but some issues need to be addressed. All comments are added directly in the manuscript. Major points include:

1. An explanation need to be added why only ERA5 was considered as data source but no alternative data sources. Are there reasons to assume that ERA5 is particularly suited for this region?
2. Statistical methods need to be explained in the methods section
3. The issue of uncertainty of rain-gauge data needs to be addressed. This includes a potential bias due to the inverse relationship between elevation and rain-gauge density.
4. The conclusion section is, in its current state, mis-named, it is a summary. “Real” conclusions should be added.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

This manuscript evaluates the performance of ERA5 reanalysis in the warm season period over the Qilian mountains. As learnd from the context, I don't think it is ready for publication in the journal Atmosphere as there are several critial issues to be addressed:

1. There are several kinds of global and regional reanalysis products except for ECMWF-series. For a peer-reviewed journal paper, it is not robust and enough to simply evaluate ERA5 in a specific region. It is difficult to understand its comprehensive performance among the global reanalysis family. Please add some other typical global reanalysis products including at least NCEP/NCAR reanalysis data, Japanese 55-year reanalysis data, NASA's MERRA2 reanalysis data in the revision.
2. The horizontal resolution in ERA5 reanalysis is 25 km, and the rain gauge observation is point-based resolution. The proximity matching method is applied in this study. It is acceptable in the plain region, but in the mountainous area, larger errors would occur as the warm-season precipitation normally happens among several squre kilometers. Please provide enough evidence to specify its shortcoming and discuss the potential biases in the Result section.
3. It seems the rain gauge data is assumed as "ground reference". It is not clear whether the rain gage observation is unbiased. It is known that wind induced undercatch is obvious in the high-cold mountains. How to quantify the reliability of the rain gauge observation in the evaluation process? Please clarify it.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors had addressed all my concerns.

Author Response

Thank you.

Reviewer 2 Report

Second review 1636090 by He et al.

It seems the authors took a “minimum effort approach” to the earlier comments. I consider revisions to be insufficient.

Concerning their replies to my major points:

1. Your explanations about the strong points of ERA5 and the advantages compared to other reanalysis products are OK – but why did you fully ignore the suggestion to add some explanatory sentences also in the manuscript? Not every reader will be fully aware of these facts
2. Revision and explanation is OK, but in your conclusion you talk about “comprehensive metrics” required to assess precipitation products. So additional statistics should be performed? Any suggestion? And why didn´t you apply them anyway?
3. It seems you missed my main point. I understand that data have undergone quality control and it is quite clear that rainfall characteristics are related to topography. Also, it is clear that the denser a rain-gauge network is, the better. However, given the inverse relationship between elevation and rain-gauge density, you not only have a better spatial representation in lower areas, you also have an increased likeliness in high-elevation areas (with less stations per area) that you miss extreme values. So the validation data you used can be expected to be biased towards characteristics of lowland areas. Improving this situation would not only require to install a denser rain-gauge network, but also to work towards a more equal spatial distribution in high- and low-elevation areas.
   Clearly, you can only work with the data you have. But a potential bias in validation data should at least be discussed, see also my earlier comments directly in the manuscript.
4. I consider the added short paragraph to be insufficient. The meaning of these conclusions is partly unclear and generally vague, see new comments directly in the manuscript.

I had mentioned in the review section that comments were added directly in the manuscript, which was uploaded. I assume the authors had access to this version (I had). So those  - mostly, but not exclusively – minor points should also be addressed during revision.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

I basically satisfied with most of the point-by-point responses in the revision. However, for the first concern, it is better to add the contexts of the associated feedback to clarify why the ERA-5 reanalysis product is merely evaluated in the Introduction section. 

In addition, the gauge data is not available from the sharing website. Please upload it with open-access using the supplemental material approach.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 3

Reviewer 2 Report

This time I am generally satisfied by the author´s revisions. Some minor points are mentioned in the attached PDF.

Authors need to carefully check mentioning and numbering of Figures in the text!

Comments for author File: Comments.pdf