the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 20 Sep 2024
Increasing water stress in Chile evidenced by novel datasets of water availability, land use and water use
Abstract. Many regions in Chile experienced an unprecedented drought from 2010 to 2022, driven by climate change and natural variability. This so-called megadrought led to severe water scarcity, causing conflicts and exposing issues in Chilean water regulations. Water-intensive agriculture in areas with limited water availability has worsened these problems, raising questions about the contributions of water extraction and climate on high water stress levels.
In this study, we evaluate water stress in Chile over the long term, from the mid-20th century to the end of the 21st century, under various climate and socio-economic scenarios. To this end, novel datasets of water availability, land use and water use were developed. Using these, we calculated the Water Stress Index (WSI) for all major basins in the country and assessed the impact of increasing water use and climate change on water stress over different time periods. Results show that most basins in semi-arid regions experienced high to extreme water stress (WSI > 40 % and WSI > 70 %, respectively) during the megadrought, mainly due to reduced water availability, but worsened by high water demand. Over time, increasing water stress in central Chile is primarily linked to rising water consumption, with a smaller contribution from water availability changes, leading to consistently high water stress levels (1990–2020 average) in several basins from Santiago northward. Under an adverse climate scenario (SSP3-7.0), megadrought-like conditions could become permanent by the end of the 21st century, with a projected 30 % drop in precipitation, resulting in high to extreme water stress in most basins in central Chile. We argue that using the WSI to assess one of the several aspects of water security offers a valuable strategy for adaptation plans. If public policy agrees on establishing quantifiable water security goals based on metrics like the WSI, different pathways of water use combined with alternative water sources can be evaluated to achieve them.
- Preprint
(4860 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on egusphere-2024-2695', Kyra Boek, 08 Nov 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2695/egusphere-2024-2695-CC1-supplement.pdf
-
RC1: 'Comment on egusphere-2024-2695', Caitlyn Hall, 06 Feb 2025
Overall Review
This paper tackles a critical issue: increasing water stress in Chile and the interplay between climate change and human water use. The integration of novel datasets, long-term historical trends, and future projections offers valuable insight into how water availability, land use, and consumption patterns have shaped the country’s current water security challenges. The study is thorough, data-driven, and policy-relevant.
However, key methodological choices need more clarity. The assumptions behind the Water Stress Index (WSI) and water balance model should be explicitly justified, particularly the exclusion of groundwater from availability estimates. Additionally, the manuscript should address uncertainties—both in historical datasets and climate model projections—more directly. The discussion on policy implications is strong but could be sharpened with concrete recommendations for decision-makers.
Despite these areas for improvement, this paper makes a strong contribution to understanding Chile’s water security trajectory. Addressing the above concerns will increase its impact and make the findings more actionable for policymakers.
Strengths
- Novel, Large-Scale Data Integration: The study compiles and analyzes long-term datasets on water availability, land use, and water use—essential for understanding historical trends and future risks.
- Robust Temporal Scope: By covering water stress from the mid-20th century through the 21st century, the paper provides a long-term perspective often missing in regional water studies.
- Future Scenario Analysis: The evaluation of different climate and socioeconomic pathways strengthens the study’s relevance to both researchers and policymakers.
- Clear Geographic Insights: The basin-level approach highlights regional disparities in water stress, making the findings useful for localized decision-making.
- Policy Relevance: The study effectively connects scientific analysis with policy discussions, emphasizing the need for quantitative water security targets in Chile.
Areas for Improvement
- Clarify Methodological Assumptions – The water balance model assumes that water availability equals precipitation minus evapotranspiration, without considering groundwater withdrawals. This may not reflect reality in highly managed basins. The authors should discuss how this limitation affects their results. Similarly, the rationale for using a WSI threshold of 40% as a stress indicator should be explicitly justified.
- Address Data Uncertainties – The study relies on multiple datasets, some of which have inherent uncertainties due to data gaps, resolution limitations, or modeling assumptions. A clear discussion of these limitations—especially regarding water use estimates—will strengthen the credibility of the findings.
- Expand Discussion of Groundwater – Groundwater overuse is a key driver of water stress in Chile, yet it receives limited attention in this study. Incorporating groundwater depletion trends or discussing how aquifer withdrawals might be masking true surface water stress would provide a more complete picture.
- Sharpen Policy Recommendations – The paper makes a strong case for integrating WSI into policy but stops short of concrete recommendations. What specific actions should be taken? How should water rights, governance structures, or infrastructure investments be adjusted in response to these findings? Explicit next steps will increase the paper’s real-world impact.
Introduction
- The introduction provides strong context for the study, linking water security to climate change and socio-economic factors.
- Page 2, Line 35-40 – "A basin is deemed to have high water stress when the Water Stress Index (WSI) –the ratio of water use to availability– exceeds 40% over the medium term (5 to 10 years)."
- The definition of WSI is useful, but the rationale for using 40% as a threshold should be explicitly stated with references to support this classification.
- Page 2, Line 45-50 – "However, obtaining accurate information can be challenging due to the poor quality and limited availability of data and ground observations in some regions."
- It would be helpful to discuss how the authors address these data gaps in their study. Are there alternative datasets or validation techniques used to mitigate uncertainties?
Methods
- Page 6, Line 90-95 – Water availability is derived from a simple water balance approach. Consider discussing whether this approach adequately captures the influence of groundwater withdrawals. Are there validation steps comparing modeled ET against observed or remote-sensing-based ET datasets?
- Page 7, Line 115-120 – Since CR2MET is a core dataset in the study, a brief discussion of its validation and potential biases would improve confidence in the results.
- Page 10, Line 175-180 – "Water availability (A) is considered here as a naturalized runoff, that is, the remaining flow from precipitation (P) and evapotranspiration (ET), without considering local disturbances."
- The assumption of no local disturbances might oversimplify real-world hydrological conditions, especially in basins with significant groundwater-surface water interactions. Consider discussing the implications of this assumption.
Results & Discussion
Page 15, Line 285-290 – "On average across continental Chile, the mean annual rates of P and ETN are estimated at 1200 mm and 430 mm, respectively, which leads to a surface water availability of approximately 770 mm yr-1."
- How do these values compare to previous studies? Providing some context (like from previous studies, especially linking to the "so what" of this article) would strengthen the reliability of these estimates.
- Page 18, Line 365-370 – "Agriculture also has a non-consumptive water use component, as some irrigation water returns to the system through infiltration and percolation."
- The study could benefit from quantifying this return flow, if possible, to better illustrate the net impact of agricultural water use.
- Page 19, Line 385-390 – The discussion on irrigation efficiency should explicitly address the paradox in which increased efficiency can lead to higher overall water use.
- Page 21, Line 425-430 – The manuscript states that "water demands approach or even exceed available surface water in a number of basins today." If groundwater use data are available, incorporating them into the discussion would strengthen the argument about unsustainable water use.
- Page 24, Line 495-500 – "Given global climate scenarios, there are only two ways to alleviate water stress in a basin: reducing consumptive water use or increasing water availability through alternative sources."
- Consider briefly mentioning the role of integrated water resource management (IWRM) and governance improvements as additional pathways to address water stress.
Conclusion & Policy Implications
- Page 26, Line 545-550 – "It should be noted that non-renewable water reserves, such as aquifers that are not in equilibrium or melting glaciers, are not considered as alternative sources to reduce the WSI."
- This is an important distinction. However, the paper could discuss the role of groundwater depletion in temporarily masking the full impact of water stress.
- Page 27, Line 560-565 – "Public policy in Chile is aware of the impacts of droughts and the challenges that climate change poses to water security."
- The discussion would benefit from more explicit recommendations on how policy can integrate WSI into national adaptation strategies.
Citation: https://doi.org/10.5194/egusphere-2024-2695-RC1 -
RC2: 'Comment on egusphere-2024-2695', Anonymous Referee #2, 03 Mar 2025
General comments
This manuscript quantifies water stress in all river basins in Chile based on a very detailled estimation of water use (U) and an appropriate high-resolution estimation of annual runoff under naturalized conditions (A) as water stress indiex WSI = U/A, for both historic conditions in the period 1960-2020 and for a potential situation in 2035-2065 under changing climate (as estimated from the median change of runoff as projected by 11 global climate models, after bias-correction of GCM climate variable output with local observations) and water use change (as simple linear extrapolation of the trends observed in the period 2000-2020). They find that in the case of low climate mitigation, precipitation at the end of the 21st century may be similarly low as during the extreme drought that occurred in the 2010s. They find that in central-northern basins of Chile, WSI values of more than 40% were reached by 2020 even though water use is regulated by water use rights. The authors suggest that public policy sets maximum WSI values that represent a threshold for water security so that then actions can be taken to not exceed WSI in each basin, taking into account future climate change.
The quantification of water availability (A) is state-of-the-art and innovative regarding spatial resolution and meteorological data used. An innovative and commendable approach taken by the manuscript is to include increased evapotranspiration due to a change in land cover since 1950 based on a detailled estimation of land use changes between 1950 and 2020 as human water use (consumptive use). It would be even better if the additional evapotranspiration caused by the artificial reservoirs would be quantified as part of the human water use U.
Unfortunately, the manuscript has a number of major weaknesses.
1) The definition of the water stress index WSI is not presented according to the literature, and it it not mentioned that the definition in the paper is not the same as in the literature. The component U generally refers to water withdrawals (= abstractions= sum of consumptive use and return flows = sum of consumptive and non-consumptive use), and not to consumptive use as seems to be the case in this manuscript (even though the specific definition is never provided, see below my comments regarding the quantification of U). Even in the publication “Alvarez-Garreton, C., Boisier, J. P., Billi, M., Lefort, I., Marinao, R., and Barría, P.: Protecting environmental flows to achieve long-term water security, J Environ Manage, 328, 116914, https://doi.org/10.1016/j.jenvman.2022.116914, 2023a.”, it is clearly stated that U are water withdrawals (and that allocated rights WUR are defined as withdrawals and thus are not comparable to consumptive water use supposed to be shown in Figure 5).
2) In various locations of the manuscript, it is wrongly stated the a WSI provides a strong quantititative basis for indicating negative impacts on watersheds (e.g. in line 499-500). In lines 39-42, the author write, without citing any publication
“There are various ways to assess water security in a territory, with methodologies varying in complexity based on the factors considered. A common and straightforward approach is to contrast water use and water availability estimates at the basin scale. A basin is deemed to have high water stress when the Water Stress Index (WSI) –the ratio of water use to availability– exceeds 40% over the medium term (5 to 10 years).”
The authors need to provide some references, and in general water availability computed over a longer time period, e.g. 30 years (see references for WSI in Alvarez-Garreton et al. 2023a). And the authors should mention that the classification of water stress occuring if WSI>0.4 (with abstractions and not consumptive use as U) is not based on evidence but a rough guess. They should clearly state from the beginning that they aim a formulating a type of WSI that is suitable for Chile or for their study, and what type of water use, water withdrawals or consumptive use they apply for WSI. I suggest that the vague term “water use” is replaced throughout the text by either consumptive use or water withdrawals (or abstractions). The term water consumption should be used only very specifically as refering to water volumes reaching households etc., but maybe is not of interest for the manuscript anyway (water consumption is not equal to consumptive use, which would have to be explained if the term consumption is used in the manuscript).
3) Usage of the term “surface water avaliability” needs to be corrected to avoid misunderstandings: A, as the difference between precipitation and actual evapotranspiration, includes groundwater recharge and thus renewable groundwater resources and the availabilty of renewable groundwater. I suggest the manuscript quantifies total availability of renewable water resources, while groundwater can supply water beyond the groundwater recharge if groundwater storage depletion occurs (with constantly falling groundwater tables). In this context, I suggest explaining shortly the situation in Chile regarding the source of water abstractions (groundwater or surface water) and any occuring groundwater depletion.
4) I have my doubts about the consistent handling of diverse types of water uses in the “CR2WU water use reconstruction”, section 2.5. On the one hand, consumptive water use for irrigation and land cover change is computed as well as non-consumptive use for irrigation. On the other hand, it appears that water withdrawals are quantfied for the other sectors (which is not explained as the term “consumption” is used starting in line 252), but consumptive uses (i.e. the part of the abstracted water that evapotranpirates during use) are not.Thus the statement in line 263 “The CR2WU dataset includes both consumptive and non-consumptive uses from LULUCF and non-LULUCF sectors.” and the entry in Table 1 is misleading. And so is the caption in for Figure 4b and the caption and title of Figure 5, where “consumptive use” appears to refer also to the drinking water, energy, manufacturing, mining and livestock water use, while these apparently refer to withdrawal rates (e.g. 145 l per cap and day for drinking water; this amount of water does not evaporate!). Either a convincing rational for mixing two types of water use (abstractions and consumptive use) is provided, or two different estimates of water use (of all sectors) are used alternatively in the WSI computation: 1) consumptive use/A 2) withdrawals/A. Both are interesting and indicate different types of stress.
5) The paper does not provide any indication how the threshold for WSI (e.g. what WSI should not be exceeded) should be determined. On what basis should stakeholder agree on such a threshold (such as 40%)? I suggest that the threshold should be be based (at least partially) on environmental flow requirements, possibly with higher eflows than today (see Alvarez-Garreton et al. 2023a).
6) Revise section 6, discussing how to achieve a (basin-specific?) threshold for WSI and also include the constraints of the study (or maybe elsewhere).
7) In Figure 4 to 8, state very clearly what water uses are included and why.
Specific comments:
In the CRWU2 files, I could only find the non-LULUCF values of water withdrawals but not the LULUCF values.
Line 45: should rely on
Line 71: due to lower
Figure 4: Explain more clearly In caption what type of water use is shown
Figure 4 b title does not seem to be correct.
Figure 4d: replace m3/s by area-specifiv values, e.g. mm/yr, as otherwise values for larger polygons are just larger because of larger area. Does it indluce hydroelectric water use? How is hydroelectric water use computed (in 4 a)?
Figure 5: Add line for population development.
Line 393: more water use by increase actual evapotranspiration (add for clarity)
Line 404: Explain Water Use Rights
Figure 6: Also show WSI for basins north of 30° S, they do have human water use.
Figure 7: Why do you show here U(cons+DW) (and not e.g. mining), in the caption it says just consumptive use?
Line 613: not clear why is a multiplier of all evapotranspiration components.
Line 619: delete “a fraction of”
Eq. B9: exp (-0.6 LAI): add minus sign
Table B2: what is “f”?
Appendix A and C: move to supplement
Citation: https://doi.org/10.5194/egusphere-2024-2695-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 535 | 284 | 274 | 1,093 | 12 | 14 |
- HTML: 535
- PDF: 284
- XML: 274
- Total: 1,093
- BibTeX: 12
- EndNote: 14
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 240 | 23 |
| Chile | 2 | 231 | 22 |
| Netherlands | 3 | 90 | 8 |
| Germany | 4 | 29 | 2 |
| United Kingdom | 5 | 28 | 2 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 240
Juan Pablo Boisier
Camila Alvarez-Garreton
Rodrigo Marinao
Mauricio Galleguillos
Our study examines water stress in Chile from mid-20th century to the end of the 21st century, using novel datasets on water availability, land use, and water use. We compute a water stress index for all basins in Chile and show that rising water use significantly contributes to water stress. We also show that a drier future is expected in central Chile and that the water stress index can be used as a tool for designing adaptation strategies.
Our study examines water stress in Chile from mid-20th century to the end of the 21st century,...