Effect of Logarithmically Transformed IMERG Precipitation Observations in WRF 4D-Var Data Assimilation System
Round 1
Reviewer 1 Report
Review Comments / Concerns
Water 825918
Effect of logarithmically transformed IMERG precipitation observations in the WRF 4D-Var data assimilation system
Summary
This manuscript sets out to address a very specific questions regarding assimilation of observed precipitation data into a forecasting model, then does that clearly and exhaustively. It is specific enough that it may not appeal to broad Water readership, but will be very helpful to those pursuing this particular area.
I have some slight editorial suggestions below that would improve readability, but beyond those recommend publication.
General questions / concerns (some of these are also reflected in Specific questions / concerns below):
- none
Specific questions / concerns
lines13-14 – suggest rewriting as “Uncertainties in precipitation estimates from numerical weather prediction (NWP) models can be reduced by assimilating precipitation observations into the NWP models.”
23-26 – sentence is confusing. Suggest you rewrite beginning with “A higher detection of heavy precipitation can be achieved with a constant error in the …:
42 – change to “However, errors in precipitation observations…”
54-55 – get rid of the “On the one had” and “On the other hand”
69-72 – it is generally thought to be poor practice to begin a sentence with a symbol. Why not just convert these periods to semi-colons?
81 – start sentence with “In eqn 2, di….”
86-87 – change to “…to avoid sub-optimal 4D-Var analysis…”
88 – “that” => “as”
108 – “chooses” => “examined”
113-114 – “…that the WRF model uses…” => “…used by the WRF model…”
119 – “(c) – (d) As in (a)-(b)…” => “Figure sets (c) – (d) and (e)-(f) are as for (a) – (b)…”
148 – “…assimilating…” => “…assimilate…”
165-168 – again, it is very confusing when you begin sentences with a symbol. It is not clear whether the period is a grammatical period or at the end of an abbreviation.
183 – “… 6-hour windows do… => “…do 6-hour windows…”
199-200 – sentence is unclear
206-209 – given that your experiment labels are not very explanatory, the captions should include a few-word description of each experiment, so that we don’t need to look back to remind ourselves
221-222 – same caption comment as above
225-225 – and yet again
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This manuscript considers the assimilation of IMERG data into a version of the WRF model, seeking to improve the precipitation estimates. The results are based on a modest dataset, but the work has merit for tackling a very challenging problem and being honest about what doesn’t seem to work as well as the title might suggest.
Major Points
- The English is pretty good; as such, I will comment on grammar and word choice.
- The Abstract should be much more specific about versions, dates, and locations.
- L.113 should (briefly) say how the interpolation is done.
- The metrics should include some kind of skill score, usually the Heidke, to help interpret the variably related lines in Fig. 6 for POD and FAR, and perhaps help declare a “winner”.
- Although Ref. 28 is the correct one for the out-of-date Version 04 data that were used in this study, I would also suggest that a more standard reference, albeit for a more modern version, might also be cited:
Huffman, G.J., D.T. Bolvin, D. Braithwaite, K. Hsu, R. Joyce, C. Kidd, E.J. Nelkin, S. Sorooshian, E.F. Stocker, J. Tan, D.B. Wolff, P. Xie, 2020: Integrated Multi-satellitE Retrievals for the Global Precipitation Measurement (GPM) mission (IMERG). Chapter 2.12 in Adv. Global Change Res., Vol. 67, Satellite Precipitation Measurement, V. Levizzani, C. Kidd, D. Kirschbaum, C. Kummerow, K. Nakamura, F.J. Turk (Ed.), Springer Nature, Dordrecht, ISBN 978-3-030-24567-2 / 978-3-030-24568-9 (eBook), 343-353. doi: /10.1007/978-3-030-24568-9_19
Minor Points
- In the Abstract and L.60, GPM needs to be written out on this first use of the acronym.
- L.42: “… However, the errors in short-interval precipitation do not”
- L.51: “… system, however, assimilating”
- L.97: It should be “Version 04” in standard GPM terminology, IMERG has already been defined, and it should be “Final Run”.
- L.125: You need a reference for this apparently standard “physics suite”.
- Fig. 7: in the titles of a) and b), and in the caption, it should be “observations used”.
Comments for author File: 
Comments.docx
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Manuscript title: Effect of logarithmically transformed IMERG precipitation observations in the WRF 4D-Var data assimilation system
Authors: Jiaying Zhang, Liao-Fan Lin, and Rafael L. Bras
Reviewer:
Summary:
This study assimilates the Integrated Multi-satellite Retrievals for GPM (IMERG) precipitation observations in the Weather Research and Forecasting (WRF) model data assimilation (WRFDA) system with a four-dimensional variational (4D-Var) method. To investigate the effect of logarithmically transformed IMERG precipitation in the WRFDA system, this study reports on several experiments with 6-hour and hourly assimilation windows, regular (non-transformed) and logarithmically transformed observations, and a constant observation error in regular and logarithmic spaces. Authors discuss the performance of rainfall forecast for each experiment. The manuscript is well written. I would like to suggest authors to add the equitable threat score (ETS) analysis to make the study more complete. The paper could be publishable in water with major revisions.
Major comments:
- The equitable threat score (ETS) is commonly used to analyze the model rainfall forecast. I would like to suggest authors to add the equitable threat score (ETS) analysis with different threshold to make the study more complete. There are two references containing ETS analysis:
Wei, C.-C. and J. Roan, 2012: Retrievals for the rainfall rate over land using Special Sensor Microwave Imager data during tropical cyclones: comparisons of scattering index, regression, and support vector regression. J. Hydrometeor., 13, 1567-2578.
Behrangi, A., K. Hsu, B. Imam, S. Sorooshian, G. J. Huffman, and R. J. Kuligowski, 2009: PERSIANN-MSA: A precipitation estimation method from satellite-based multispectral analysis. J. Hydrometeor., 10, 1414-1429.
Minor comments:
- Line 106: Any reason for the consideration of constant error of 0.3 mm/hr?
- In Fig. 3 and 4, authors are suggested plot Fig. 3 and 4 with upper limit values of X- and Y-axis.
- Line 108-116: Please give more information about three summer convective events over the central United States (e.g. the mechanism for the development of heavy rainfall; the similarity and differences among these three events).
- Authors should provide the explanation for green colored circles filled with red and blue color in Fig.4.
- Line 177: “Im contrast”: change to “In contrast”
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 3 Report
Manuscript title: Effect of logarithmically transformed IMERG precipitation observations in the WRF 4D-Var data assimilation system
Authors: Jiaying Zhang, Liao-Fan Lin, and Rafael L. Bras
Reviewer:
Summary:
This study assimilates the Integrated Multi-satellite Retrievals for GPM (IMERG) precipitation observations in the Weather Research and Forecasting (WRF) model data assimilation (WRFDA) system with a four-dimensional variational (4D-Var) method. To investigate the effect of logarithmically transformed IMERG precipitation in the WRFDA system, this study reports on several experiments with 6-hour and hourly assimilation windows, regular (non-transformed) and logarithmically transformed observations, and a constant observation error in regular and logarithmic spaces. Authors discuss the performance of rainfall forecast for each experiment. The manuscript is improved and well written. The paper could be publishable in water with minor revisions.
Minor comments:
- Line 199-201 & Table 1: “Figure 5 shows the statistical metrics (the threshold for POD, FAR, and ETS is 0.001 mm/hr) of experiments of non-transformed/logarithmically transformed observations with a constant observation error in regular space (EXP2/EXP3) or in the logarithmic space (EXP4/EXP5).” Please double check all descriptions in Table 1.
Comments for author File: Comments.pdf
Author Response
We added the errors in the logarithmic space for EXP2 and EXP4 in Table 1. Now it is clearer that the observation errors are constant in regular space for EXP2 and EXP3 and are constant in the logarithmic space for EXP4 and EXP5. Thanks.
Author Response File: Author Response.pdf
