An improved global remote-sensing-based surface soil moisture (RSSSM) dataset covering 2003–2018

Chen, Yongzhe; Feng, Xiaoming; Fu, Bojie

doi:https://doi.org/10.5194/essd-13-1-2021

Articles | Volume 13, issue 1

https://doi.org/10.5194/essd-13-1-2021

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

https://doi.org/10.5194/essd-13-1-2021

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

Articles | Volume 13, issue 1

Data description paper

|

05 Jan 2021

Data description paper |

| 05 Jan 2021

An improved global remote-sensing-based surface soil moisture (RSSSM) dataset covering 2003–2018

Yongzhe Chen, Xiaoming Feng, and Bojie Fu

Download

Final revised paper (published on 05 Jan 2021)
Supplement to the final revised paper
Preprint (discussion started on 08 May 2020)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review comments for essd-2020-59 "A new dataset of satellite observation-based global surface soil moisture covering 2003–2018 " by Chen, Y., Feng, X., and Fu, B.', Anonymous Referee #1, 09 Jun 2020
- AC1: 'Author comment #1', Yongzhe Chen, 14 Aug 2020
RC2: 'Comments on “A new dataset of satellite observation-based global surface soil moisture covering 2003~2018”', Anonymous Referee #2, 29 Jun 2020
- AC2: 'Author comment #2', Yongzhe Chen, 14 Aug 2020
RC3: 'A new dataset of satellite observation-based global surface soil moisture covering 2003–2018', Anonymous Referee #3, 24 Jul 2020
- AC3: 'Author comment #3', Yongzhe Chen, 14 Aug 2020

Peer-review completion

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

AR by Yongzhe Chen on behalf of the Authors (10 Sep 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (22 Sep 2020) by Giulio G.R. Iovine

RR by Anonymous Referee #4 (07 Oct 2020)

Suggestions for revision or reasons for rejection

Review of
An improved dataset of satellite observation-based global surface soil moisture covering 2003~2018 (RSSSM)
by Chen et al.

This is a very valuable contribution to the research on long-term soil moisture data. The approach is technically sound and state-of-the-art.
In the introduction the authors too strongly blame existing soil moisture products without discussing their advantages and drawbacks in detail. Also other methods than NN to generate long-term time series such as Copulas are not mentioned. Introduction needs a much clearer structure.
Also the description of the neural net approach is not very clear. Figure 2 helps, but the full iterative and localized approach is still not clear. Also the separation into monthly ~10-day bins is not justified. Why not using a strict 10 day temporal resolution? During validation the scale difference between coarse resolution of most satellite SSM products as compared to point-scale in situ measurements is not discussed. Also the discussion is very descriptive focusing on the statistics, but I would like to see deeper interpretations by linking environmental characteristics, microwave observation methods and NN predictions.
The language needs to be improved. All abbreviations need to be introduced first. The authors should speak about surface soil moisture, not surface moisture or similar.
I am missing also a description of the data set itself, i.e. which format, auxiliary data etc.
In general, I recommend a minor revision of the manuscript.

Specific comments:
L. 8f: be more specific, what is lacking?
L. 12: elaborate?
L. 11ff: strange formulation. Iterative and localized?
L. 16: introduce RSSSM
L. 19f: why in cold and arid regions?
L. 21f: which period? Is it valid to use the data set for trend analysis?
L. 33: where do the uncertainties come from?
L. 47f: explain abbreviations
L. 47ff: change to AMSR-E, also in the following
L. 51ff: usually is wrong, reformulate. Reduced performance due to complex topography or dense vegetation
L. 56: reference for SMOS is Kerr et al. (2001)
L. 57: better than shorter wavelengths
L. 61: reformulate: …and incorporated hardware RFI mitigation
L. 64f: published a long-term surface soil moisture dataset under the Climate Change Initiative (CCI). Delete: or Essential Climate Variable (ECV)
L. 71: justify low quality, why?
L. 72: why is the merging algorithm probably too simple? Justify!
L. 77ff: it is not true that temporal changes are mainly driven by model simulations, reformulate
L. 85-90: it is not clear what the authors want to say here
L. 92: polarized reflectivity? Isn’t it emission?
L. 103: this data has been
L. 115: iterative 5-step neural network
L. 123ff: It has to be noted that the SMAP_E grid is 9km only, but that the spatial resolution of that product is around ~20km. The 9km is misleading and should be clarified also to correctly interpret the final product of this study.
L. 126: The nominal penetration depth
L. 134: ASCAT soil water index
L. 135: not developed by Copernicus, but by EUMETSAT. It is provided by the ESA Copernicus Land Monitoring Service
L. 135f: why was the SMAP porosity used, and not the porosity provided with the ASCAT product?
140ff: Highly questionable to use X-band for soil moisture retrieval, maybe that is the reason why the results of this study show improvements for cold and arid regions with low vegetation. C-band RFI is known to be high over the US, but also over the rest of the globe?
L. 150: again, why not C-band also for Windsat?
L. 169f: not all products are retrieved by a RTM. Additionally, typical all retrieval methods use a vegetation and/or a LST information. Don´t you add here those characteristics twice? Please discuss.
L. 182: sand and clay fractions can be named soil texture
L. 189f: the WARP change detection algorithm is applied to ASCAT after using the different angles to remove the vegetation contribution. Please check the ASCAT SM retrieval and modify accordingly.
L. 211: that
L. 252ff: please justify this filtering.
L. 323f: it is not clear why a NN sorting is necessary
L. 335: the Russian networks do not have any data in this period.
L. 406f: is that an independent comparison? If not, please indicate.
L. 434f: The ASCAT problem in arid areas is known and might be related to changes in soil scattering where the change detection method assumptions are not longer valid.

Hide

RR by Anonymous Referee #1 (11 Oct 2020)

Suggestions for revision or reasons for rejection

I've got the impression that the authors have sufficiently addressed my previous comments and questions. I support the publication of this data description paper after the following minor points had been considered:

L134: "The first is ASCAT" change to something like: "The first satellite soil moisture product", otherwise it reads as if it was a quality impact factor product.

L180: You speak of 9 quality impact factors but you only list 8 and in the overview Fig. 1 there are only 7 different types distinguished.

L193: GEOV2-LAI is missing in the overview figure (Fig. 1)

L211: that that

L223, L224, continue the enumeration of quality impact factors as you started with "first, second, third, ..." until you reach 9th as defined in the introduction

L249: Still required to state the Toolbox used in Matlab 2016 and the name of the neural network training function

L339: which as a -> which has a

L395: "grid" please change all occurrences to "grid cell". A grid refers to an aggregate of cells, so a 0.5° x 0.5° unit corresponds to a grid cell.

L546: if you write "most" you should state which ones are better than yours

L573: You could name the Cosmic-Ray Neutron Sensing method (CRNS) and the COSMSOS (http://cosmos.hwr.arizona.edu/) network here as a potential provider of root zone integrated large scale soil moisture with global coverage

Andreasen, M., Jensen, K.H., Desilets, D., Franz, T.E., Zreda, M., Bogena, H.R. and Looms, M.C. (2017), Status and Perspectives on the Cosmic‐Ray Neutron Method for Soil Moisture Estimation and Other Environmental Science Applications. Vadose Zone Journal, 16: 1-11 vzj2017.04.0086. doi:10.2136/vzj2017.04.0086

L990: Figure caption: "Overview of the periods of the different soil moisture datasets..."

L995: Figure caption: "Flow chart for the ..."

L1002: provide time range for (c) RSSSM vs site measured soil Moisture

Table 2 & 3: you could highlight the best performing dataset (per column) by setting their values bold

Language: the writing has improved since the initial version of the manuscript but some parts are still difficult to grasp at first reading.

Hide

ED: Reconsider after major revisions (16 Oct 2020) by Giulio G.R. Iovine

AR by Yongzhe Chen on behalf of the Authors (13 Nov 2020) Author's response

ED: Publish as is (18 Nov 2020) by Giulio G.R. Iovine

AR by Yongzhe Chen on behalf of the Authors (19 Nov 2020) Manuscript

Download

Article (18531 KB)
Full-text XML

Short summary

Soil moisture can greatly influence the ecosystem but is hard to monitor at the global scale. By calibrating and combining 11 different products derived from satellite observation, we developed a new global surface soil moisture dataset spanning from 2003 to 2018 with high accuracy. Using this new dataset, not only can the global long-term trends be derived, but also the seasonal variation and spatial distribution of surface soil moisture at different latitudes can be better studied.