Technical note: A hydrological routing scheme for the Ecosystem Demography model (ED2+R) tested in the Tapajós River basin in the Brazilian Amazon

Pereira, Fabio F.; Farinosi, Fabio; Arias, Mauricio E.; Lee, Eunjee; Briscoe, John; Moorcroft, Paul R.

doi:https://doi.org/10.5194/hess-21-4629-2017

Articles | Volume 21, issue 9

https://doi.org/10.5194/hess-21-4629-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/hess-21-4629-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 21, issue 9

Technical note

|

14 Sep 2017

Technical note |

| 14 Sep 2017

Technical note: A hydrological routing scheme for the Ecosystem Demography model (ED2+R) tested in the Tapajós River basin in the Brazilian Amazon

Fabio F. Pereira, Fabio Farinosi, Mauricio E. Arias, Eunjee Lee, John Briscoe, and Paul R. Moorcroft

Download

Final revised paper (published on 14 Sep 2017)
Preprint (discussion started on 04 May 2016)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'Revision of 'Technical Note: A hydrological routing scheme for the 2 Ecosystem Demography model (ED2+R), Hydrol. Earth Syst. Sci. Discuss., doi:10.5194/hess-2016-114, 2016'', Anonymous Referee #1, 18 Jun 2016
- AC1: 'Author's responses to Referee #1', Fabio Farinosi, 31 Jul 2016
RC2: 'Referee Report', Anonymous Referee #2, 19 Jun 2016
- AC2: 'Author's responses to Referee #2', Fabio Farinosi, 31 Jul 2016

Peer-review completion

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

ED: Reconsider after major revisions (16 Aug 2016) by Graham Jewitt

AR by Fabio Farinosi on behalf of the Authors (31 Jan 2017) Manuscript

ED: Referee Nomination & Report Request started (01 Mar 2017) by Graham Jewitt

RR by Hartley Bulcock (27 Mar 2017)

RR by Anonymous Referee #3 (10 May 2017)

RR by Anonymous Referee #4 (23 May 2017)

Suggestions for revision or reasons for rejection

Overall:
I found this a very informative technical note on adding a hydrological routing scheme to a terrestrial biosphere or land-surface model and felt the paper is well written. There are minor typographic errors and some methodological details that require additional specificity to avoid confusion (particularly the terminology used for the residence time adjustment parameters). Specific suggestions are given below.

For further testing of this model structure beyond this methodological technical note, I would advocate a more in-depth calibration procedure and description (how parameter ranges and parameter sets were selected for trial, etc) with more analysis of resulting parameter values between the subcatchments. I would also like to see the methodology trialed in different bioclimatic regions.

Specific edits and suggestions:

Title – consider adding something to specify that the method was applied and tested to a sample catchment in the Amazon Basin (suggests need to trial further in other regions)

P3 L10 typo: should be “…does not allow us to properly…” or, alternatively, “…does not allow for proper simulation of the…”
P4 L32 suggest starting a new paragraph from the sentence: “In this technical note….”
P5 L3 colloquial English suggestion: “…, thereby reproducing daily river flows…”
P5 L6 typo: should be “…changes such as…”
P5 L9 typo: should be “…used as a computational…”
P9 L17 What weighting factors were used for the KGE? Please specify
P12 L7 Specify the number of model runs used and the values selected for the other hydrological parameters (CS, CI, CB, etc) during this step of the calibration.

P12 L16-17 (& L30, etc) If I understand correctly CS, CI, and CB are not residence times, they are residence time adjustment coefficients according to Collischonn et al. (2007). They adjust an initial estimate of residence time which is based on the topographic relief of the grid cell and the grid cell size (hence flow-path length). This should be specified in the text and the C parameters should not be referred to as residence times in text, tables, or figure captions, but rather as adjustment coefficients or something along those lines. The way it is currently written, referring to them as residence times, the reader is left confused about why the different subcatchments have such extremely different CS, CI, and CB values thinking they are similar to the more typical residence time parameter in a linear reservoir equation. The comparative values of these parameters for surface, interflow, and baseflow reservoirs for some subcatchments are also somewhat surprising: there are a few where the values are quite similar when one would expect baseflow residence times to be significantly longer than shallower reservoirs. Showing the calculated average residence times (calculated using the C parameters and topography) for the different compartments for the grid cells in each subcatchment added in a table or modifying Figure 7 would be more physically meaningful for someone looking at the conceptualization of hydrological processes.

P15 Discussion section – in this section there is discussion of using smaller grid cell sizes with higher resolution input data in order to improve the accuracy of the simulation. It should be noted when saying this that the ED2+R model does not route groundwater between cells, it only routes the groundwater that leaves an area via river baseflow. Not considering groundwater flow between grid cells is a more reasonable simplification when grid cells are very large, but if they were significantly reduced groundwater routing between grid cells may need to be added to maintain or improve realism and accuracy. At finer spatial scales not all model grid cells would be expected to contain channels that drain the subsurface reservoirs being explicitly considered.

P15 L25 colloquial English suggestion: “…too shallow to realistically…”
P16 L15 typo: should be “…performance of the model by providing…”
P16 L31 typo: should be “…typical of these kinds of models.”

P29 L3-4 caption for Figure 3 (c) should not be “land use” as land usage has not been mapped or specified in the image. Instead consider, “(c) aerial imagery of Tapajos river basin illustrating land cover diversity in the catchment….” Or change the figure to include a map of the land use that was used in the model (cite Hurtts et al 2006)

P30 L12-14 caption for Figure 6 – reword and insert the parameter abbreviation to add clarity: “Model sensitivity to the initial conditions of baseflow (QB)…”

P32 L1-6 caption for Figure 7 – Specify that the C parameters are adjustment coefficients of residence time (and not actually residence times).
Consider remaking this figure instead using the average residence time value for the grid cells in a subcatchment for the surface, interflow, and baseflow reservoir. This would still be a representation of the parameter value impact for the C parameters because the remaining values used to calculate residence time in Collischonn et al. (2007) are derived from the DEM and grid cell size and so are fixed per grid cell and not calibrated. Showing the residence time value would be more easy to analyze from a hydrologic process conceptualization point of view and to compare the processes simulated for the different subcatchments with different geologies and geomorphologies (the axes could be kept consistent between subcatchments). Otherwise the C parameter values on their own seem rather arbitrary and their values unexpected (comparing the surface flow, interflow, and baseflow values within and between catchments). This could be a very interesting and informative figure if reformatted in that way.
Also specify that the colour bar below each diagram is indicative of the model Nash-Sutcliffe Equilibrium (NS) for the model realization for the given set of parameter values. In previous diagrams and text NSE is used rather than NS – one abbreviation should be used throughout.

P34 L2 Table 1 caption – Are these the calibrated parameter values? If so specify this in the caption. Again don’t refer to the C values as residence times, rather call them residence time adjustment factors or something along those lines. Otherwise the values are confusing for those used to looking at residence times and linear reservoir equations.

Hide

ED: Publish subject to minor revisions (further review by Editor) (12 Jun 2017) by Graham Jewitt

AR by Fabio Farinosi on behalf of the Authors (20 Jul 2017) Author's response Manuscript

ED: Publish subject to technical corrections (02 Aug 2017) by Graham Jewitt

AR by Fabio Farinosi on behalf of the Authors (08 Aug 2017) Author's response Manuscript

Short summary

ED2 is a terrestrial biosphere model (TBM) suited for investigating combined impacts of changes in climate, atmospheric CO₂, and land cover on the water cycle. In this study, we describe the integration of ED2 with a hydrological routing scheme. The resulting ED2+R model calculates the lateral propagation of surface and subsurface runoff resulting from the TBM and determines spatiotemporal patterns of river flows. We successfully evaluated the ED2+R model in the Tapajós, Brazilian Amazon.