Evaluation of retrieval methods for planetary boundary layer height based on radiosonde data

Li, Hui; Liu, Boming; Ma, Xin; Jin, Shikuan; Ma, Yingying; Zhao, Yuefeng; Gong, Wei

doi:https://doi.org/10.5194/amt-14-5977-2021

Articles | Volume 14, issue 9

https://doi.org/10.5194/amt-14-5977-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/amt-14-5977-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 14, issue 9

Research article

|

03 Sep 2021

Research article |

| 03 Sep 2021

Evaluation of retrieval methods for planetary boundary layer height based on radiosonde data

Hui Li, Boming Liu, Xin Ma, Shikuan Jin, Yingying Ma, Yuefeng Zhao, and Wei Gong

Download

Final revised paper (published on 03 Sep 2021)
Preprint (discussion started on 11 May 2021)

Interactive discussion

Status: closed

RC1:
'Comment on amt-2021-118', Anonymous Referee #1, 27 May 2021

Comments on the manuscript entitled " Evaluation of retrieval methods for planetary boundary layer height based on radiosonde data" by Li et al., submitted to Atmospheric Measurement Techniques (AMT-2021-118):

The manuscript described and inter-compared performances of four methods (PM, RM, GM_RH, and GM_θ) which were widely used to estimate PBLH from RS data. In general, the study fit the scope of the journal and the manuscript was well organized. These results shown in the manuscript can be regarded as a useful reference when selecting boundary layer algorithms. Additionally, there are still several points that need to be clarified before it could be considered for acceptance.

Specific comments:

1) In Abstract section, some sentences are not clear enough between Line 19 and Line 25 of Page 1, such as “PBLH from PM is the lowest under all and SBL classifications, and the highest under CBL and NBL classifications”. Please rephrase or clarify.

2) The study focused on estimating the performances of four PBLH calculation methods. But, in the Introduction section, the description of the advance of the subject (namely, the various comparison and estimation of PBLH methods in existing research) is not sufficient enough.

3) How much data were used in the study obtained from sites of Beijing, Wuhan, Changsha at 0600 UTC? Since the 0600 UTC is afternoon at local, there should be more CBL and NBL cases in these three cities (as shown in Figure 1 and Page 6 Line 1-2).

4) In addition to the TIL, have any additional indicators been added to filter data in order to remove cases under extreme weather conditions? Will the extreme weather, such as rain, snow, fog and storms, impact on the estimation of the boundary layer for RS data?

5) Whether the performance estimations and the OP method are affected by geographic location? I noticed that the mean value of PBLH obtained by OP method are lowest or nearly lowest among the four methods in some cities, such as 57494, 57687, 59758, and 59948.

6) It is better to add a flow chart for the OP method as described in section of 3.4.

Citation: https://doi.org/10.5194/amt-2021-118-RC1
- RC2:
  'Reply on RC1', Anonymous Referee #2, 12 Jun 2021
  
  Comments on the manuscript entitled " Evaluation of retrieval methods for planetary
  
  boundary layer height based on radiosonde data" by Li et al., submitted to Atmospheric
  
  Measurement Techniques (AMT-2021-118):
  
  This study focuses on the performance of four common RS methods under different thermodynamic stability conditions and proposes an optimal processing flow for the RS data retrieval of PBLH. This study provides an optimal RS standard value retrieval method for further inversion of the PBLH through artificial intelligence algorithms. The article has a clear structure and contributes to this area. But in my opinion, there are still some problems should be solved before publishing.
  
  Major concerns:
  
  For gradient method, the author mentioned that the threshold values of the potential temperature and RH vertical gradients were set as 0.003 K/m and 0/m, respectively. Why need to set this threshold? Similarly, why was the threshold of Ri set to 0.25, and whether the sensitivity test has been carried out?
  
  About the RS data, why did author choose only nine sites for the experiment? Will there be similar results using data from all sites in the country?
  
  Specific concerns:
  
  Section 1: In the third part of the introduction, only a few methods compared in this paper are described in the description of the existing RS data retrieval methods of PBLH, which should be described more comprehensively.
  
  P4-L18-19: This article studies PBLH retrieval method based on RS data. What is the role of radar wind profiler here?
  
  P6-L6: Although there is an explanation for the abbreviation of GM_θ in the abstract, what does the θ refers to here, should be explained again.
  
  P6-L19: Does the “rib” here refer to the Richardson number? If so, “b” should be in the form of subscript. And the specific meaning of ri_b needs to be explained.
  
  Figure 2: Why is the proportion of various categories not 100% at All time? At different times, the proportion of different categories is relative to all the cases of a site, or relative to the effective cases?
  
  Section 3.3: This section analyzes the consistency of different algorithms under various classification conditions, and finds out the reasons for the inconsistency. However, we notice that the inconsistency ratio under SBL classification is higher than that under other classification conditions. It is suggested to make a key explanation for the high inconsistency ratio under SBL classification.
  
  Citation: https://doi.org/10.5194/amt-2021-118-RC2
  - AC2: 'Reply on RC2', Boming Liu, 22 Jul 2021
    
    Please check the attachment!
    
    Citation: https://doi.org/10.5194/amt-2021-118-AC2
- AC1: 'Reply on RC1', Boming Liu, 22 Jul 2021
  
  Please check the attachment!
  
  Citation: https://doi.org/10.5194/amt-2021-118-AC1

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Boming Liu on behalf of the Authors (22 Jul 2021) Author's response Manuscript

ED: Publish as is (02 Aug 2021) by Cheng Liu

Short summary

Radiosonde (RS) is widely used to detect the vertical structures of the planetary boundary layer (PBL), and numerous methods have been proposed for retrieving PBL height (PBLH) from RS data. However, an algorithm that is suitable under all atmospheric conditions does not exist. This study evaluates the performance of four common PBLH algorithms under different thermodynamic stability conditions based on RS data.