This paper has created the most comprehensive long-term historical monthly climate dataset using Guidelines on Best Practices for Climate Data Rescue. This quality-controlled dataset is an invaluable resource to the scientific community and I see that it has been made openly available on the Zenodo data repository and the SMHI data web portal. The dataset and corresponding information is also openly shared and will be very useful for climate reconstructions and historical wind analysis. I would like to commend the authors on their hard work in producing this dataset. I would most certainly recommend this article for publication.

Suggestion and minor comments:

 Line 235: For the future work you could look at engaging with local universities to introduce data rescue into the classroom and encourage undergraduate students to get involved in digitising small amounts of data but  “many hands make light work”. Students can digitise large volumes of meteorological observations quickly and effectively. We have just successful integrated a project into the 2^nd Undergraduates Geography curriculum at Maynooth University Ireland. Students helped digitise some unique historical Sub-Saharan African climate data from imaged sheets. Ryan et al 2021 also previously ran a successful data rescue project with undergraduate students.

Line 105: How many of these stations were able to be joined to contemporary station records providing longer extended records? Was this not included in your criteria for digitisation? Also did the archive sheets contain any other variables? If so are there any plans to digitise these data?

Line 195: Do you have any metrics on the scale of qc issues found in the digitised data? What percentage of missing data was present in each station after qc was conducted?

Reference

Ryan, C., Duffy, C., Broderick, C., Thorne, P. W., Curley, M., Walsh, S., Daly, C., Treanor, M., & Murphy, C. (2018). Integrating Data Rescue into the Classroom, Bulletin of the American Meteorological Society, 99(9), 1757-1764. Retrieved Aug 31, 2021, from https://journals.ametsoc.org/view/journals/bams/99/9/bams-d-17-0147.1.xm

correction 

first line of review should read; 

"This paper has created a long-term historical wind dataset using Guidelines on Best Practices for Climate Data Rescue"

This work has added new and valuable digitized instrumental measurements of wind speed and direction covering 1920s-1940s, to some extent, many observations only stored in paper journals have been rescued. It is meaningful for the researchers to well understand wind changes and multidecadal variability enveloping one full century. And this work has also made some contributions to the international data rescue. However, there are still some shortcomings in this paper need to be improved.

Specific comments:

1. L183: It is recommended to delete “section 4.2.1 Scanning and digitization” or incorporate this section into section 3.2 or 3.3, simply describe the contents of 4.2.1 in section 3.2 or 3.3.

1. In the whole paper, the section of homogenizing wind data has not been introduced, and some key techniques such as establishing reference series, detecting and adjusting shifts are also missing. The data connection provided in this paper is not enough to support the complete data processing for wind observations.

1. The article is to support the publication of the wind dataset, so it should display some outstanding results in the processing of quality control and homogenization, rather than a reference Zhou et al. (2022) mentioned in “section 7 conclusions” (L240) can represent. It is recommended to analyze some contents which can be presented by tables or figures: (1) number of suspected error data before and after quality control, cause statistics, and how to deal with them, etc. ; (2) the series of wind observations raw and adjusted; (3) assessment of the climate characteristics of the newly wind data, etc.

1. L219: It is recommended to add some explanations on the shortcomings and improvements in this work in “section 5 discussion”.

1. L236: It is recommended to incorporate “section 6 outlook” into “section 5 discussion”.

Minor comments:

Please adjust or correct the grammar or format in the following sentences.

1. L46: "stilling" and "reversal" periods is due is the low availability of centennial wind series.

1. L82: This created a pressure drop in the tube moving a pointer that showed the wind speed on a scale.

1. L160-L162: Observations have been continuously registered since at Malmslätt since 1945. Before that storing of journals were more random which caused the data gap from July 1936 to January 1939.

1. L165-L166: The data gap in the 1940s is due to observations that by mistake were not digitized in the project.

1. L220: This work contributes to satisfy the need as stated by (WMO, 2016) to increase.

1. L222: the wind climate in Sweden and globally as highlighted by (IPCC, 2021).

1. L237-L238: At SMHI the work to rescue and digitize historical meteorological, hydrological and oceanographic observations from the archives of paper journals spanning over more than 150 years continues and.