HydroBlocks v0.2: enabling a field-scale two-way coupling between the land surface and river networks in Earth system models

Chaney, Nathaniel W.; Torres-Rojas, Laura; Vergopolan, Noemi; Fisher, Colby K.

doi:https://doi.org/10.5194/gmd-14-6813-2021

Articles | Volume 14, issue 11

https://doi.org/10.5194/gmd-14-6813-2021

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

https://doi.org/10.5194/gmd-14-6813-2021

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

Articles | Volume 14, issue 11

Development and technical paper

|

09 Nov 2021

Development and technical paper |

| 09 Nov 2021

HydroBlocks v0.2: enabling a field-scale two-way coupling between the land surface and river networks in Earth system models

Nathaniel W. Chaney, Laura Torres-Rojas, Noemi Vergopolan, and Colby K. Fisher

Download

Final revised paper (published on 09 Nov 2021)
Preprint (discussion started on 28 Oct 2020)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

SC1: 'executive editor comment on gmd-2020-291', Astrid Kerkweg, 14 Nov 2020
RC1: 'Comments on GMD-2020-291', Anonymous Referee #1, 17 Nov 2020
RC2: 'Review comments', Dai Yamazaki, 18 Nov 2020
AC1: 'Comment on gmd-2020-291', Nathaniel Chaney, 21 Jan 2021

Peer-review completion

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

AR by Nathaniel Chaney on behalf of the Authors (03 Jun 2021) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (10 Jun 2021) by Jeffrey Neal

RR by Anonymous Referee #1 (27 Jun 2021)

Suggestions for revision or reasons for rejection

I appreciate the authors' responses, which addressed part of my comments. I still would like to challenge the authors to further elevate this manuscript. In particular, I have the following feedback on the authors' responses.

It is true that most ESMs don't represent two-way coupling between their land and river components. A complete representation of it involves both the suitable process description and, more importantly, a proper parameterization strategy. The former is relatively easy since the ESM modelers just need to borrow some governing equations from the well-established fields of hydrology or hydraulics. The latter, however, is much more challenging and, in my opinion, more critical in a model development paper like this. For a new process description in ESMs, a good parameterization usually means the parameter values are provided via some a priori estimation or extensive calibrations and ready to use for the potential users. As we are all aware, for ESMs it is usually not practical to expect the users to define/calibrate the parameter values by themselves because 1) it is computationally prohibitive to perform sufficient parameter calibration with ESMs like we used to do with much cheaper hydrological models and 2) very often the ESMs users have a very diversified background and many of them do not have adequate hydrology background to carry out such parameter estimation on their own. In a nutshell, when adding a new process representation in ESMs, it is most important to demonstrate that its corresponding parameterization strategy is compatible with the process description and hence effectively improves the model predictions in some aspects. In some sense, without a proper parameterization, adding a new process description will not necessarily bring better model predictions, particularly for sophisticated models like ESMs. Therefore, it is not obviously beyond the scope of this study to add sufficient model evaluations against the ARM SGP observations that the authors already have, and equally importantly, other observations at a regional or global scale where ESMs are typically applied at. In any case, it is my understanding that, when publishing a new process development in ESMs, it is more important to have a good parameterization strategy and adequately validate the new development against some observations. If the editors feel that it is not necessarily the standard for GMD, I am also fine with it.

Moreover, the explanation of using a constant and uniform velocity for the HRU-level impulse response function is not very convincing. The authors stated that "Since the fixed flow velocity assumes the flow to the channel is not impacted by more/less water down the hillslope, this assumption will be valid." This assumption "the flow to the channel is not impacted by more/less water down the hillslope" is not quite rigorous. At an hourly or shorter time step, this assumption is certainly not correct, because overland flow velocity and the travel time will be affected by the surface runoff depth, as indicated by the well-known kinematic wave equations, e.g., Manning's equation. At a daily time step, such a statement may be ok, if slightly rephrased, since the impacts of runoff depth are somehow negligible in terms of travel time, but then all the other advantages of having two-way land-river interactions and sub-grid heterogeneities might not be significant at a daily or longer time step as well.

Hide

ED: Reconsider after major revisions (07 Jul 2021) by Jeffrey Neal

AR by Nathaniel Chaney on behalf of the Authors (14 Sep 2021) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (23 Sep 2021) by Jeffrey Neal

Dear Nathaniel,
Thank you for your revision to the paper. I'm sorry this has taken a while for me to handle. Given the difference of opinion between the reviewer and yourself regarding the need to include a parametrisation for the model I wanted to review the paper again myself before making a decision.
Firstly, I'm in agreement with the reviewer that the parameterisation of this model type is of critical importance and that in many ways the design of the model is easier because the governing equations are well established. Nevertheless, I appreciate that identifying suitable observations, test cases and calibration strategy for this is difficult and requires significant space/effort to set out properly. Furthermore, if I understand correctly, the bulk fluxes are simple not sensitive enough to the coupling to aid calibration thus details spatial data is essential.
The discussion between the reviewer and yourself has convinced me that this is really half a model description with the parametrisation approach to come later. Given the complexity of the model I think a standalone description is worth publishing and that GMD is the appropriate journal for that. However, since the relative performance of the new model is not well understood this should be made very clear to the reader. This is done at the start of the discussion “it remains unclear if this paramerterization improves the macroscale modelling of surface fluxes and inundation…” But I think a similar statement should be made either in the aims, conclusions, or both. Such that it’s clear that further research is necessary before the model description is complete.
This is a trivial change so I will accept subject to consideration of my point above. Please also feel free to explain if I’ve misunderstood something because my expertise is in the inundation modelling components of this work rather than the LSM.
Best wishes,
Jeff

Hide

AR by Nathaniel Chaney on behalf of the Authors (05 Oct 2021) Author's response Manuscript

Short summary

Although there have been significant advances in river routing and sub-grid heterogeneity (i.e., tiling) schemes in Earth system models over the past decades, there has yet to be a concerted effort to couple these two concepts. This paper aims to bridge this gap through the development of a two-way coupling between tiling schemes and river networks in the HydroBlocks land surface model. The scheme is implemented and tested over a 1 arc degree domain in Oklahoma, United States.