The manuscript submitted by Boitard et al. derived an AGB dataset from SMOS L-VOD using a method developed by Rodriguez-Fernandez et al. (2018) and Mialon et al. (2020) based on a logistic fit between the AGB and VOD relationship. Compared to the mentioned studies, the manuscript evaluates the impact of the descending/ascending overpasses and regional models on the AGB retrieval. Overall, the manuscript reads well and fits nicely into the scope of the ESSD journal. Furthermore, an additional AGB dataset that provides more than a single static map could benefit the carbon and ecosystem monitoring community. However, the authors need to address the following comments:

Major comments:

1. A definition of AGB needs to be added to the introduction. Does it refer solely to forest above-ground biomass? Furthermore, significant work regarding the derivation of AGB from microwave remote sensing needs to be included, e.g., using brightness temperature. Additionally, specific properties of L-VOD are unfairly attributed only to SMOS L-VOD, while significant literature has been put out on using SMAP L-VOD for similar applications.
2. The authors need to explain how the reference maps were aggregated to the SMOS spatial resolution. The aggregation method might be important, particularly in the case of CCI AGB, which comes in a resolution of 100m x 100 m. How were the uncertainties (standard deviations) used in the context of aggregating to the SMOS resolution?
3. The methods chapter is not very clear.
   • The authors mention that all years of SMOS data and other AGB reference maps were used for the final dataset production. However, the paper describes only the methodology applied for 2018, a year where both CCI AGB and SMOS L-VOD estimates are available. How did the authors proceed with the years with no available reference maps?  
   • Were all CCI AGB maps used? CCI AGB provides change maps with a quality flag concerning the reliability of AGB change. Where these flags used?
   • How was the Avitabile map used? Is there any bias between CCI AGB and Avitabile AGB? Furthermore, Avitabile corresponds to 2010, while Fig. 3 shows yearly AGB estimates starting in 2011. For what year was Avitabile used?
4. The evaluation is very minimal, particularly in the temporal domain. Given that the study provides a dynamic AGB dataset, the plausibility of temporal patterns is paramount. The authors should consider validating change patterns, e.g., per land cover.
5. The few time series provided do not agree very well with either CCI AGB (Fig. 8) or Xu AGB (Fig.9). While the authors mention the disagreements in the discussion chapter, further explanations are needed.

Minor comments:

Line 34: "satellite remote sensing of AGB" - direct AGB measurements do not exist - change to deriving AGB from satellite remote sensing.

Lines 55-57: The mentioned properties are not inherent only to SMOS L-VOD. Move above to the general properties of L-VOD.

Fig. 3: Step 4 "Validation" - Change "different parameters" to, e.g., "different model fits/factors".

Fig. 3: Are the AGB estimates starting in 2011, not 2010?

Line 384: 2011, not 2010?

Abbreviations are not consistent between plots and captions (CCI-2018-v4; CCI AGB reprojected)

Fig 10: SMOS L-VOD 2018?

This paper describes a dataset of above ground biomass (AGB) obtained from satellite remote sensing. The L-Band Vegetation Optical Depth (VOD) from the Soil Moisture and Ocean Salinity (SMOS) mission linked to existing AGB maps allows defining spatial relationships between L-VOD and AGB. AGB is then determined for 2011-2023 from the L-VOD data. In this study, the impacts of the orbit type (using only ascending or descending data or combining both) and of using a global spatial calibration vs. several regional calibrations of L-VOD – AGB relationships were determined.

The new AGB dataset is then compared against the ESA CCI and Xu et al. (2021) AGB datasets. When using the single global calibration, the L-VOD AGB dataset has lower AGB in equatorial Africa than the ESA CCI AGB dataset, as the logistic fit is not able to reach AGB values over 300 Mg ha-1, which are present in parts of Africa, but not in other densely vegetated regions such as the Amazon rainforest. AGB estimates in equatorial Africa can be improved with the regional calibrations.

General comments:

I believe that the study is of interest to the community and that another above ground biomass dataset covering a longer time period will be useful. The paper is generally well organized, however, in a few sections its clarity could be improved.

The authors should explain the validity and usefulness of comparing an AGB dataset with the Xu et al. (2021) dataset, which combines above ground and below ground biomass more clearly. Why was this dataset chosen, if it covers a different variable?

I suggest that the above and below comments should be addressed before publication.

Specific comments:

- l. 29: Clearly define AGB and whether it includes biomass of all vegetation or for example just forests.

- l. 98: Explain what exactly version 700 of the data is.

- l. 104: “located under high” should be “located in the high”

- Figure 1 caption: “The colored rectangles show the extent of the different regions considered in the study” -> Maybe mention a bit more here about what these regions are or refer to where in the manuscript they are explained.

- l. 112: “Santoro and Cartus (2023)” -> Maybe, refer to it as the ESA CCI (or CCI) map like you do below.

- l. 125: “resampled to EASE 2 grid” -> Should be “resampled to the EASE 2 grid”

- ll. 144-145: Why do you compare the total living biomass from Xu et al. with the CCI AGB? As you’re comparing total living biomass with AGB, if the total living biomass exceeds the CCI AGB, it doesn't tell you whether the difference between the two datasets is just in the BGB or the AGB as well.

- Figure 2: Why is there no panel b? Why not label panel c as b, even if it is in the second row?

- l. 147: “derive to AGB” -> Remove “to” or replace with “the”.