A seawater desalination scheme for global hydrological models

Hanasaki, Naota; Yoshikawa, Sayaka; Kakinuma, Kaoru; Kanae, Shinjiro

doi:https://doi.org/10.5194/hess-20-4143-2016

Articles | Volume 20, issue 10

https://doi.org/10.5194/hess-20-4143-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/hess-20-4143-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 20, issue 10

Research article

|

12 Oct 2016

Research article |

| 12 Oct 2016

A seawater desalination scheme for global hydrological models

Naota Hanasaki, Sayaka Yoshikawa, Kaoru Kakinuma, and Shinjiro Kanae

Download

Final revised paper (published on 12 Oct 2016)
Supplement to the final revised paper
Preprint (discussion started on 31 Mar 2016)
Supplement to the preprint

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

SC1: 'Efforts in addressing the following comments would be appreciated', Yaling Liu, 01 Apr 2016
- AC2: 'Response to Dr. Yaling Liu', Naota Hanasaki, 21 Jun 2016
RC1: 'review', Anonymous Referee #1, 01 Apr 2016
- AC3: 'Response to Referee 1', Naota Hanasaki, 21 Jun 2016
RC2: '"A seawater desalination scheme for global hydrological models” by N. Hanasaki et al. - Review', Anonymous Referee #2, 02 Apr 2016
- AC4: 'Response to Referee 2', Naota Hanasaki, 21 Jun 2016
EC1: 'editor comment', Murugesu Sivapalan, 27 May 2016
- AC1: 'Response to Editor', Naota Hanasaki, 21 Jun 2016

Peer-review completion

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

ED: Reconsider after major revisions (25 Jun 2016) by Murugesu Sivapalan

AR by Naota Hanasaki on behalf of the Authors (27 Jun 2016) Author's response Manuscript

ED: Referee Nomination & Report Request started (01 Jul 2016) by Murugesu Sivapalan

RR by Anonymous Referee #2 (06 Jul 2016)

RR by Anonymous Referee #1 (09 Jul 2016)

Suggestions for revision or reasons for rejection

Although the authors have tried to address the concerns that were raised during the first round of review, I am still not satisfied with the quality and the scientific contribution of this paper to warrant its publication in HESS. In a nutshell, the paper relies on a large dataset of desalination plants (DesalData) to identify grid cells that rely on desalinated water through a set of predetermined conditions (GDPC>14k, aridity<7.5%, distance<~165km) , and then through a set of assumptions, the model establishes the amount of desalinated water that would need to be produced. The established thresholds are kept the same globally and over time and as such they could yield strong implications about the future dependence on desalinated water. For example, the per capita income threshold of 14k does not account for the depreciation value of money over time, and since GDPC is expected to increase under all the 5 SSPs, this assumption will lead to wide spread of desal under all the SSP scenarios even the defragmented (and relatively poor) world of SSP3. As ground water resources start to deplete, there is no guarantee that the 165km threshold will hold and this is already seen in the case of the city of Riyadh, which is about 400km from the nearest coast. The same can be said about the aridity index threshold. Also the fact that conditions/threshold/assumptions were based on the same data that the authors are using to validate the model defeats that purpose. One would expect there is an independent dataset for calibration and another one for validation, which was not the case for this study. Even when basing the validation on the actual year (2005) to derive the grid areas depending on desalinated water and their production estimates, there are many exceptions as discussed by the authors (e.g., Spain and Israel). Finally the paper can still improve from proofreading to clarify the writing. I have tried to highlight some examples below.

Line 18: change ‘period’ to ‘year’

Lines 25-26: I would suggest omitting the second to last sentence. If you keep it, I would reduce the number of significant figures used or use billion $ unit instead of millions (1.1-10.5 & 1.5-22.1 Billion USD). Although I am not familiar with how much desal plants typically cost, but the numbers do look small. Is it only the capital investment cost that you have accounted for? or also the levelized cost over the lifetime of the plant including operation and maintenance, and the energy cost often associated with such activities, plus piping structures, pumping and distribution costs?

Line 32: ‘there is a growing concern over the sustainability of’

Line 45: Add a more recent citation beside (Bremere et al., 2001).

Line 47: the introduction lacks any information about the current capacity & production of desal and even the historical evolution beyond citing figure S1. For example, when citing that Kim et al came up with the estimate of 250km3/yr by 2100, it is hard for the reader to give sense to how big or small that number is in comparison to current use.

Line 91: give the numbers for the rest of categories, ‘followed by brackish water, saline inland and river water, and others.’ If I am understanding this correctly, the 13.3 is the largest portion of the 21.9 total, and the other categories add up to the remaining 8.6km3, right?

Line 93: “First, we selected the desalination plants (hereafter Major Plants) to be included in our analyses.” Are not you missing the word ‘major’ in that sentence?

Lines 93-94: “The detailed selection criteria and the rationale is shown in Table S1 and Supplemental Text.” Table S1 lists the criteria but does not say anything about the rationale. For example, it is still not clear what was the rationale for using only 613 large of 17000+ plants available? The SM does not really address this point as claimed by the authors. The only piece that relates to this aspect is the following “Finally, we selected plants of larger than 10000m3d-1 of capacity. It corresponds to approximately 1.5mm yr-1 of water resources assuming the area of a grid cell is 2500km2(0.5°×0.5°), which is same order of the magnitude of other water resources components simulated in H08.”

Lines 95-96: “For example, DesalData contains some records indicating that plants in mountainous inland regions use seawater as the source, which is erroneous. The plants excluded from this study are listed in Table S2with reasons.” The mountainous example is not listed as a reason for any of the 11 excluded plants in table S2. Does that mean that criterion was never used to exclude any plant? If so, then replace with the criteria mentioned in table S2.

Lines 103-107: it is not clear yet in the flow of the paper why there is a need to collect data on the distance of desal plants and major cities. Briefly explain

Line 149: By inputting ..." rephrase the sentence. replace 'inputting' with 'using', 'extracts' with 'identifies', and 'that' with 'where'. also specify what gridded global maps are used as inputs. Also what conditions? The sentence needs to be rephrased.

Line 153: delete ‘latest’

Lines 155-157: replace “but it is not necessarily true that desalination plants are located” with ‘but desalination plants are not necessarily located’

Lines 162-164: the three conditions stand out as subjective at best. It seems that it would make more sense to discuss the data analysis first to justify the selection of these criteria

Line 167: delete “Conditions B, C, A correspond to each.”

Lines 174-176: again these seem random assumptions without sufficient justification to their logic. Where is the 30-80% coming from? The tendency to promise info to come later in the manuscript makes the manuscript harder to follow.

Line 168: “data used and spatial resolution of interest, hence modified if SDM is applied in a different simulation settings.” – not sure what this really means. Rephrase.

Line 206: replace ‘was installed’ with ‘were installed’

Line 245: “We approximated it to 30-80% and set as Assumption C.” why not use 28-77% as shown by the data?

Figure S3: Are you accounting for the diminishing purchasing power of money over time (discount rate) when accounting for the 14000$/person threshold? The SSP3 should be a ‘poor’ world with high population and difficulties to both mitigate and adapt.

Table S2: what does it mean that ‘lat/long data collapsed’? this does not make sense to me. Are you saying these points (lat/long) fall outside of the country boundary (Spain in this case)? Or something else.

Hide

RR by Anonymous Referee #3 (12 Jul 2016)

Suggestions for revision or reasons for rejection

The authors of the revised manuscript present a model for areas and volumes of use of desalinated seawater (SDM) that is proposed to be suitable for incorporation into global hydrological models (GHMs) to estimate current seawater use (2005) but in particular to derive scenarios of future use of desalinated water. They also generated three scenarios for the time periods 2011-2040 and 2041-2070 based on the global Shared Socioeconomic Pathways (SSP) scenarios of future population and GDP.

In my opinion, the capacity of the developed SDM for simulating current/historical use of desalinated seawater is so poor such that it is not meaningful or sensible to include the model in GHMs. SDM predicts the existence of desalinated seawater use only as a function of aridity, GPD and distance to ocean. Table 2, a comparison of Fig. 1 and Fig. 4 as well as Fig. 6 show that the assumption that desalinated seawater is used only in very arid grid cell with an aridity index of 0.075 or less precludes the simulation of existing seawater use in Spain and Israel as well as in coastal areas with desalination plants in Libya, Australia, the USA, India and China even in the future (see comp. Fig. 6e and f to 6a-d). While with this aridity criterion, the current bulk of desalinated seawater that is produced on the hyperarid Arabian peninsula is captured, this does not make the model suitable for simulating the extension of seawater desalination to other less arid coastal areas that may be expected (and has been observed) with increasing water stress and decreasing cost of desalination. Unfortunately, the authors do not describe well how they derived their model and how the results using alternative predictors would be. I think that predictors with a higher predictive capacity should be tested. If it is not possible to derive a more “predictive” model than it is better to refrain from simulating future changes in the use of desalinated seawater. This is the main reason why I think the manuscript should be rejected.

In addition, the three derived scenarios are driven by climate (change), GDP change and domestic and industrial water use change only, with future new grid cells with desalinated seawater use mainly being determined by the different GDP developments in the three scenarios. In the spirit of scenario development, it would have been interesting to represent the different storylines of the SSPs (sustainability vs. middle of the road vs. fragmentation) by e.g. favoring waste water treatment over desalination in the sustainability scenarios SSP1.

Finally, the presentation of the research even in the revised version is still only fair. For example, neither in the abstract nor in the introduction is it mentioned that they only simulate desalinated sea water use for domestic and industrial uses. The relationship between aridity index and the existence of desalination plants which is central to the model is not shown while a global map of the aridity index is (lines 214-217). It is not discussed what can be learned from then new panels Fig. 6e and 6f. There are many typos, and the manuscript would require language editing.

Hide

ED: Reconsider after major revisions (14 Jul 2016) by Murugesu Sivapalan

AR by Naota Hanasaki on behalf of the Authors (25 Aug 2016) Author's response Manuscript

ED: Publish as is (16 Sep 2016) by Murugesu Sivapalan

Short summary

Although seawater desalination has been widely implemented and used as a key source of water in arid regions, it has been seldom included in global water resource assessments based on numerical simulations. We first developed a global model to estimate the areal extent and production of seawater desalination which was designed to be incorporated with global hydrological models. The model was applied to future periods under three distinct socioeconomic conditions.