Statistical post-processing of reanalysis wind speeds at hub heights using a diagnostic wind model and neural networks

Brune, Sebastian; Keller, Jan D.

doi:https://doi.org/10.5194/wes-7-1905-2022

Articles | Volume 7, issue 5

https://doi.org/10.5194/wes-7-1905-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/wes-7-1905-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 7, issue 5

Research article

|

16 Sep 2022

Research article |

| 16 Sep 2022

Statistical post-processing of reanalysis wind speeds at hub heights using a diagnostic wind model and neural networks

Sebastian Brune and Jan D. Keller

Download

Final revised paper (published on 16 Sep 2022)
Preprint (discussion started on 14 Mar 2022)

Interactive discussion

Status: closed

RC1: 'Comment on wes-2022-20', Michael Mifsud, 12 Apr 2022

Line 48: Could benefit from a literature review, following which the aim of the research isd described, rather than being part of the intooduction.

lines 88 - 124 could benefit with a description of symbols used.

Section 3.2

Why ANNs, how do they compare against other Machine Learning models? Why this specific ANN model?

Figure 2: Can we see a terrain map near the charts?

Figure 3: Assuming RMSE is Root Mean Square Error, how is this being calculated? Why RSME? Why is it better than other metrics such as MAE?

What are you comparing in FiG 3?

Figures 4&5: Mainly unclear, you mention colours which are not visible. I have difficulty understanding it.

Line 203: Why 60%, 20% 20%? what advantages? is it comapred to 70%, 15% 15%?

Assume that NN means neural networks, work could benefit from a list of abbreviations. diffficult to follow all abbreviations.

Citation: https://doi.org/10.5194/wes-2022-20-RC1
RC2: 'Comment on wes-2022-20', Anonymous Referee #2, 24 Jun 2022

Excellent paper, it was a pleaure to read

three minor comments:

page 4: delete footnote 6, table 2 is included

page 10 line 173: COSMO instead of COMSO

page 11 line 195: COSMO instead of COMSO

a review of the revised manuscript is not necessary

Citation: https://doi.org/10.5194/wes-2022-20-RC2
AC1: 'Comment on wes-2022-20', Sebastian Brune, 21 Jul 2022

Thanks for the suggestions to both Reviewer. We added a file with a detailed response to all comments.

Citation: https://doi.org/10.5194/wes-2022-20-AC1

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Sebastian Brune on behalf of the Authors (21 Jul 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (21 Jul 2022) by Rebecca Barthelmie

RR by Michael Mifsud (11 Aug 2022)

ED: Publish as is (12 Aug 2022) by Rebecca Barthelmie

ED: Publish as is (12 Aug 2022) by Jakob Mann (Chief editor)

AR by Sebastian Brune on behalf of the Authors (29 Aug 2022)

Short summary

A post-processing of the wind speed of the regional reanalysis COSMO-REA6 in Central Europe is performed based on a combined physical and statistical approach. The physical basis is provided by downscaling wind speeds with the help of a diagnostic wind model, which reduces the horizontal grid point spacing by a factor of 8. The statistical correction using a neural network based on different variables of the reanalysis leads to an improvement of 30 % in RMSE compared to COSMO-REA6.