Towards ice-thickness inversion: an evaluation of global digital elevation models (DEMs) in the glacierized Tibetan Plateau

Chen, Wenfeng; Yao, Tandong; Zhang, Guoqing; Li, Fei; Zheng, Guoxiong; Zhou, Yushan; Xu, Fenglin

doi:https://doi.org/10.5194/tc-16-197-2022

Articles | Volume 16, issue 1

https://doi.org/10.5194/tc-16-197-2022

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

https://doi.org/10.5194/tc-16-197-2022

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

Articles | Volume 16, issue 1

Research article

|

21 Jan 2022

Research article |

| 21 Jan 2022

Towards ice-thickness inversion: an evaluation of global digital elevation models (DEMs) in the glacierized Tibetan Plateau

Wenfeng Chen, Tandong Yao, Guoqing Zhang, Fei Li, Guoxiong Zheng, Yushan Zhou, and Fenglin Xu

Download

Final revised paper (published on 21 Jan 2022)
Supplement to the final revised paper
Preprint (discussion started on 13 Jul 2021)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Comment on tc-2021-197', Anonymous Referee #1, 10 Aug 2021

My detailed comments on the manuscript can be found in the attached annotated pdf. The manuscript discusses a relevant topic (accuracy of different DEMs on the Tibetan Plateau and impact on ice thickness reconstruction) and overall I find that the methods are robust and useful conclusions are drawn. I do have quite a few comments (see attachment), most prominently related to 1) the way differences are calculated [I suggest to subtract the ICESat-2 data from the DEM datasets rather than the other way around; this would avoid a lot of confusion], 2) the large number of figures [some suggestions for removing figures are given], 3) clarification or explanation [in many places details about the methods and presented figures and tables are missing; additionally, numerous textual revisions may help to clarify the content], 4) transferability of results to other regions [some discussion/speculation on how the presented findings on DEM and ice thickness comparison may translate to other mountainous regions would be interesting to add].

Citation: https://doi.org/10.5194/tc-2021-197-RC1
- AC1: 'Reply on RC1', Wenfeng Chen, 30 Aug 2021
  
  Many thanks to the reviewer for the critical comments. We have carefully addressed all the comments. 1) And we have changed the way differences are calculated, Tables and Figures are all updated. 2) We have deleted two figures and simplified two figures. 3) Details about the method and figures are added. 4) Further discussions are also given in the manuscript. Our point-by-point responses are attached below in blue, while the original reviewers’ comments are in black. The detailed response and revision in the manuscript could be found attached.
  
  Citation: https://doi.org/10.5194/tc-2021-197-AC1
CC1:
'Comment on tc-2021-197: vertical datum offsets for global DEMs', David Shean, 10 Aug 2021

This is a brief comment based on a quick skim of the discussion paper. It looks like a nice contribution, but I believe there is a fundamental issue with the methodology involving global DEM preparation.

I noticed what looks like a systamatic vertical datum offset in some of the results (e.g., Figure 12), with offsets of ~30 m between ICESat-2 elevation (height above ellipsoid) and a subset of the global DEM elevations (height above respective geoid model - see documentation for specific geoid used by each DEM). It's important to account for this offset before doing any analysis. The apparent 30 m bias is due to the geoid offset, and is not representative of the quality of the DEM. This bias will propagate to RMSE numbers, which will impact the conclusions.
It should be relatively straightforward to correct each of the input DEMs to provide ellipsoidal heights using available offset grids (https://cdn.proj.org/), and repeat the analysis. I can't remember if the RGI-TOPO dataset (https://rgitools.readthedocs.io/en/latest/dems.html) has already accounted for vertical datum differences between different sources. One can also use the OpenTopography GlobalDEM API (https://portal.opentopography.org/apidocs/#/Public/getGlobalDem), which provides versions of some global DEMs with ellipsoidal heights rather than the orthometric heights.

Citation: https://doi.org/10.5194/tc-2021-197-CC1
- CC2: 'Reply on CC1', Wenfeng Chen, 30 Aug 2021
  
  We thank Dr David Shean for pointing out this and the information provided. These were very valuable. The offset of ~30 m between ICESat-2 elevation and the global DEM elevations is indeed due to the differences of datum reference. We ignored the reference differences between various DEMs and ICESat-2. In fact, ICESat-2 data, NASADEM_SHHPv001 and TanDEM-X are based on WGS84 ellipsoid reference, and the other four DEMs are all based on EGM96 geoid (Table 1). We have added this information to Table 1. We updated the data and the geoidheight function provided by MATLAB was used to calculate geoid height to unify their references. All the analyses are repeated. All the figures and Tables are updated thoroughly. Our main conclusion doesn’t change, we still concluded that NASADEM performed best and would be the best choice for ice-thickness estimates over the TP. The details about the revisions could be found attached.
  
  Citation: https://doi.org/10.5194/tc-2021-197-CC2
RC2:
'Comment on tc-2021-197', Anonymous Referee #2, 03 Oct 2021

The comment was uploaded in the form of a supplement: https://tc.copernicus.org/preprints/tc-2021-197/tc-2021-197-RC2-supplement.pdf

Citation: https://doi.org/10.5194/tc-2021-197-RC2
- AC2: 'Reply on RC2', Wenfeng Chen, 22 Oct 2021
  
  We are grateful to the anonymous reviewer for the constructive comments on our manuscript. We have carefully addressed all the comments below. The response was uploaded in the form of a supplement
  
  Citation: https://doi.org/10.5194/tc-2021-197-AC2

Peer review completion

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

ED: Reconsider after major revisions (further review by editor and referees) (08 Nov 2021) by Arjen Stroeven

AR by Wenfeng Chen on behalf of the Authors (17 Nov 2021) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (29 Nov 2021) by Arjen Stroeven

RR by Anonymous Referee #2 (13 Dec 2021)

RR by Anonymous Referee #1 (15 Dec 2021)

ED: Publish subject to minor revisions (review by editor) (18 Dec 2021) by Arjen Stroeven

AR by Guoqing Zhang on behalf of the Authors (21 Dec 2021) Author's response Author's tracked changes Manuscript

ED: Publish as is (21 Dec 2021) by Arjen Stroeven

AR by Guoqing Zhang on behalf of the Authors (21 Dec 2021) Manuscript

Short summary

A digital elevation model (DEM) is a prerequisite for estimating regional glacier thickness. Our study first compared six widely used global DEMs over the glacierized Tibetan Plateau by using ICESat-2 (Ice, Cloud and land Elevation Satellite) laser altimetry data. Our results show that NASADEM had the best accuracy. We conclude that NASADEM would be the best choice for ice-thickness estimation over the Tibetan Plateau through an intercomparison of four ice-thickness inversion models.