Dear Authors and Colleagues
Thank you for your contributions, thanks for the authors for the replies to the reviews and for the revised version of the manuscript.
The published global time series on belowground autotrophic respiration is of importance. As the data description and data publication do not yet fulfill the requirements of ESSD and improvements are needed a major revision of the manuscript and a minor revision and new publication of the dataset is needed.

The manuscript provides a lot of detailed information and an interesting cross comparison with the Hashimoto et al. 2015 global data set on belowground respiration.
However the description of the data set, its data sources and the data generation is lost in the complexity of the paper. This happens because the article is not exclusively focused on the dataset but put also a focus on its application. Recommendations to the authors: for ESSD the focus of the manuscript should be on data sets and products. the word ‘study’ is frequently used and could sometimes be exchanged with ‘data set’

Requirements for publication:
1) data sources need to be cited correctly

I) It is not clear which version of the Global data base was used: in the Introduction it is cited as “from the most updated global soil respiration dataset (SRDB v4, doi: 10.5194/bg-7-1915-2010, Bond-Lamberty and Thomson, 2010a) “– a formally better citation would be: Global Database of Soil Respiration Data Version 4 (SRDB v4, Bond-Lamberty and Thomson, 2010a) However the manuscript states on p.18, L407… this study used an up-to-date field observational database developed from SRDB up to the end 2018. Does this refer to this current version (2018)? Bond-Lamberty, and Thomson. 2018. A Global Database of Soil Respiration Data, Version 4.0. ORNL DAAC, Oak Ridge, Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/1578

In chapter 2.1, p.5 L121, details on “Q10”, “Q11”, “Q12”, “Q13” and “Q14 should be given

II) the gridded input data sets are not correctly referenced

Examples: p.6, L136 “soil nitrogen content was downloaded from Spatial Data Access Tool (https://webmap.ornl.gov/ogc/index.jsp)“ is referring to a web visualization and download service of this data set. However, the information needed for this input variable is it’s correct naming and the correct citation of Authors, or a Program. In this case the variable name is ‘Granule soil total nitrogen density’, the data set collection: ‘Global Soil Data Task’, the Program ‘International Geosphere-Biosphere Programme (IGBP)’, the year of the first publication of this dataset: 2000.
A DOI was assigned in 2014 that should be shown in the reference list.

The authors should provide in 2.2. a table with more details as it is now displayed in the supplement, with correct references and units (e.g., weight% of total nitrogen).
Most probably all the gridded time series data used as input in this study are constrained onto the time period 1980 to 2012?

all abbreviations needed to be spelled out when they are used the first time in the text: e.g. CRU TS, NOAA/ESRL etc.

2) Suggestion 2.6 to provide a more specific subtitle e.g. ‘cross comparison with Hashimoto 2015 Soil Respiration’
The authors need to provide details on the Hashimoto et al 2015 product, a short overview on key points how it was produced (to avoid that the readers of this manuscript need to read Hashimoto et al. 2015 to understand the data set production and pelicularities) and how the data set was treated for further analyses.
Example on the details of input data and the methods that Hashimoto et al 2015 used for the generation of their product are required, e.g, that gridded air temperature and precipitation data were taken from East Anglia Climatic Research Unit CRU, soil respiration data from SRDB, Bond-Lamberty and Thomson, 2010
How is the product autotrophic respiration from the Hashimoto et al 2015 product derived (Hashimoto et al product representing the total soil respiration) – this retrieval step is so far not described as a processing that it is undertaken by the author team.
the product from Hashimoto et al 2015 should always be named together with the publication year, or shorter for Figures etc. as Hashimoto 2015. The naming of the Random Forest driven data set as ‘data driven product’ is still a very general description, one could construct a name using the Random Forest method, and also take this name instead of referring in figures to ‘this study’.

3) For publication in ESSD the authors need to optimize the metadata of their product and publish the new version on figshare.

I) -global attributes: should contain a meta data field with the name of the product;
-variable: the long_name should contain the name of the variable, e.g. autotrophic respiration, the unit should contain ‘C m-2 y-1’;
-z: the long_name should contain more details on z, the unit should contain ‘years since 1980’

file name is now Ra.10Fold.CV.720.360.1980.2012. Consider to change the file name to a less technically-focused file name, consider more explaining keywords, e.g. Respiration_autotroph_glob-author_year.

II) Abstract and title on figshare
Now it reads ‘Global variability of belowground autotrophic respiration in terrestrial ecosystems’ ..the netcdf file represent the global belowground autotrophic respiration from 1980 to 2012 with a resolution of half degree. The unit is g C m-2 a-1 the product was produced by global published observations with the linkage of global climate soil and other environmental variables using random forest.’
Suggestion to change the title more towards a product (it does not need to be the same title as in ESSD)
e.g. title: Global belowground autotrophic respiration from 1980 to 2012
abstract text – suggestion to put more technical details in, some suggestions what information in the figshare abstract can be provided below
global belowground autotrophic respiration from 1980 to 2012 …. Global Database of Soil Respiration Data (Bond-Lamberty and Thomson…), …environmental climate and soil variables … random forest. …The unit is C m-2 y-1’. The spatial resolution is 0.5 degree latitude longitude on WGS84 ellipsoid (?), temporal resolution yearly …

4) chapter 5 data availability
Describe here details of the data set taken from the figshare abstract, e.g. that it is a global data set with x bands the variable yearly autotrophic respiration unit C m-2 y-1’

I look forward to your final manuscript,
Best wishes,
Birgit Heim

Dear author team

we are very pleased to accept your revised manuscript and revised data publication and data presentation in ESSD.

please consider, prior to publication, the 2 technical comments from reviewer 1
before uploading your final version

The reviewers rated your publication in the second round as good and very significant.
we are looking forward to publish your paper

best regards, Birgit Heim