Locality-based 3-D multiple-point statistics reconstruction  using 2-D geological cross sections

Chen, Qiyu; Mariethoz, Gregoire; Liu, Gang; Comunian, Alessandro; Ma, Xiaogang

doi:https://doi.org/10.5194/hess-22-6547-2018

Articles | Volume 22, issue 12

https://doi.org/10.5194/hess-22-6547-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-22-6547-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 22, issue 12

Research article

|

20 Dec 2018

Research article |

| 20 Dec 2018

Locality-based 3-D multiple-point statistics reconstruction using 2-D geological cross sections

Qiyu Chen, Gregoire Mariethoz, Gang Liu, Alessandro Comunian, and Xiaogang Ma

Download

Final revised paper (published on 20 Dec 2018)
Preprint (discussion started on 14 Jun 2018)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'On the lack of three-dimensional training images', Anonymous Referee #1, 24 Jun 2018
- AC1: 'Response to Reviewer #1', Gang Liu, 23 Sep 2018
RC2: 'Review of "Locality-based 3D multiple-point statistics reconstruction using 2D geological cross-sections"', Anonymous Referee #2, 31 Jul 2018
- AC2: 'Response to Reviewer #2', Gang Liu, 23 Sep 2018
RC3: 'Interactive comment on “Locality-based 3-D multiple-point statistics reconstruction using 2-D geological cross-sections” by Qiyu Chen et al.', Anonymous Referee #3, 25 Aug 2018
- AC3: 'Response to Reviewer #3', Gang Liu, 23 Sep 2018

Peer-review completion

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

ED: Publish subject to revisions (further review by editor and referees) (03 Oct 2018) by Philippe Ackerer

ED: Publish subject to revisions (further review by editor and referees) (04 Oct 2018) by Philippe Ackerer

AR by Gang Liu on behalf of the Authors (05 Oct 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (10 Oct 2018) by Philippe Ackerer

RR by Thomas Hermans (16 Oct 2018)

Suggestions for revision or reasons for rejection

Dear authors,

The revised version of your manuscript "Locality-based 3-D multiple-point statistics reconstruction using 2-D geological cross-sections" is strongly improved compared to the initial submission. You answered thoroughly my comments and I think the paper is ready for publication. I only have a few remarks that could help to clarify some points of the paper:

1) Have you considered giving a name to your approach? It would help removing the occurrence in the text of "our approach" along with DS and S2Dcd.
2) P3L4: remote sensing and geophysical images ?
3) P8L8-20: I still don't understand the explanation of the aggregation method. Maybe I wrongly understand your sentences. You first state that log-linear pooling is used to combine probabilities with significant correlation. Then you say that two parallel sub-sections often contains similar pattern (to me, this corresponds to the idea of correlation) and that you thus use an additive aggregation operator. From my point of view, there is a contradiction here.
4) P9L26-27. If I understand correctly, the data event is not really divided on several grids. If you have many points available close to the node to simulate, there will still be considered in the 1st multi-grid and the radius will be smaller through the maximum number of neighbours. The chance it happens is just low.
5) P10L1-3. I would explicitly state that out of plane hard data will be disregarded.
6) P10. Algorithm 1 - Point 4. Your process is fully random, i.e. that the random path is in the 3D space and not within a specific section (from what I understand). I think it is worth mentioning.
7) P15. Table 2. It is probably out of the scope of the paper, but it seems that the number of geobodies cannot be reproduced using directional 2D images. Do you think it would be better with a 3D image or is it rather related to the stochastic simulation process?
8) P22L11-13. I guess you would get a stronger discrimination if the width of the Kernel was reduced. I am a little bit surprised that DS still gets 29% of the contribution.

Hide

RR by Kashif Mahmud (12 Nov 2018)

ED: Publish subject to minor revisions (review by editor) (22 Nov 2018) by Philippe Ackerer

AR by Gang Liu on behalf of the Authors (26 Nov 2018) Author's response Manuscript

ED: Publish as is (06 Dec 2018) by Philippe Ackerer

Short summary

One of the critical issues in MPS simulation is the difficulty in obtaining a credible 3-D training image. We propose an MPS-based 3-D reconstruction method on the basis of 2-D cross sections, making 3-D training images unnecessary. The main advantages of this approach are the high computational efficiency and a relaxation of the stationarity assumption. The results, in comparison with previous MPS methods, show better performance in portraying anisotropy characteristics and in CPU cost.