the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
A system dynamic model to quantify the impacts of water resources allocation on water-energy-food-society (WEFS) nexus
- 1State Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan 430072, China
- 2Hubei Province Key Lab of Water System Science for Sponge City Construction, Wuhan University, Wuhan 430072, China
- 1State Key Laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan 430072, China
- 2Hubei Province Key Lab of Water System Science for Sponge City Construction, Wuhan University, Wuhan 430072, China
Abstract. Sustainable management of water-energy-food (WEF) nexus remains an urgent challenge, as interactions between WEF and community sensitivity and reservoir operation in water system are often neglected. This paper aims to provide a new approach for modeling WEF nexus by incorporating community sensitivity and reservoirs operation into the system. The co-evolution behaviors of the nexus across water, energy, food and society (WEFS) were simulated by the system dynamic model. The reservoirs operation was simulated to determine water supply for energy and food systems by the Interactive River-Aquifer Simulation water resources allocations model. Shortage rates for water, energy and food resulted from the simulations were used to qualify their impacts on WEFS nexus through environmental awareness (EA) in society system. Community sensitivity indicated by EA can adjust the co-evolution behaviors of WEFS nexus through feedback loops. The proposed approach was applied to the mid-lower reaches of Hanjiang river basin in China as a case study. Results show that EA accumulation is mainly from shortages of water and energy, and the available water and energy are the vital resources to sustain WEFS nexus. Feedback driven by EA effectively keeps the system from collapsing and contributes to the concordant development of WEFS nexus. Water resources allocation can remarkably ensure water supply through reservoirs operation, decreasing water shortage rate from 16.60 % to 7.53 %. The resource constraining the WEFS nexus is transferred from water to energy. This paper therefore contributes to the understanding of interactions across WEFS system and helps the efficiency improving of resources management.
Yujie Zeng et al.
Status: final response (author comments only)
-
RC1: 'Comment on hess-2021-521', Anonymous Referee #1, 04 Nov 2021
This study addresses the phenomenon that the water, energy, and food crises that human society is facing are highly interconnected issues, and their evolutions would further stimulate human response actions, which would in turn (re)shape the evolution trajectories of the FEW systems. In doing so, the authors develop a holistic sociohydrologic model, which not only mimics the water, energy, and food systems but the related human components (e.g., population, GDP, industry, agriculture) are also incorporated endogenously. Overall, the work is interesting and represents a very important direction for extending the scope of sociohydrology, which has been discussed particularly by Di Baldassarre et, al, Sociohydrology: Scientific Challenges in Addressing the Sustainable Development Goals https://doi.org/10.1029/2018WR023901. In this sense, I think this manuscript is a valuable contribution to the scientific progress within the scope of sociohydrology. However, I do have some concerns and suggestions that need to be addressed, which are listed below.
- The text and grammar should be revised throughout. There are many places (too many to be listed) where the language is unclear and misleading.
- I suggest the authors give a more detailed description of figure 1. This figure is very important for understanding the overall feedback relationships between the model variables. Currently, I am not very clear about the feedback relationships.
- I have some concerns about equations (2)-(5). First, this seems not the Malthus growth model. In the Malthus growth model, the right side of equations 2-5 should be N, G, A, and WQ, respectively, instead of N0, etc. please check if it is a typo. Second, there is an exponential term which the authors call the technology effect, dampening the growth rate of the state variables. This is not very convincing. I believe that technology development would contribute to water conservation activities and thus reduce water use quota, but I do not understand why it would have a negative effect on GDP, population and crop area, this is somewhat counter-intuitive. Third, equation (5). Why is there a negative sign in front of WQ? From table 2, rqwu is already a negative value (i.e., -0.02). If you intend to indicate that the water use quota is decreasing over time, one negative sign needs to be removed. In addition, in this case, the exponential term would dampen the decreasing rate of water use quota. This might not be reasonable, because technology development is always supposed to accelerate the decreasing of water use quota instead of dampening it. Fourth, there is a term representing the effect of GDP on water use quota in equation (5). I assume the rationale is that GDP development would prompt the advancement of water-saving technology. But the effect of technology has already been considered by the exponential term. I think perhaps the equation (5) is over-complex. Fifth, line 155, the authors claim that this study considers municipal and rural water consumption, industrial water consumption and agricultural water consumption, so I think there should be a distinction of water use quota for each of these types of water use. However, there seems no distinction between the different types of water use in equation (5).
- The description of the water resources allocation in section 2.1.2 is too simple. I cannot understand the rationale behind equations 6 and 7. Especially, reservoir operation is an important focus of this study, I suggest the authors give some more detailed descriptions of the water resources allocation processes. Currently, it is difficult to see how the water shortage rate is calculated in equation 7.
- Equation 8 has the same problem as equation 5, please see comment (3).
- I am a bit confused about how energy consumption is defined in this study. In equation 9, energy consumption is calculated by multiplying water supply by energy use quota, so I assume that energy use quota is defined as the energy demand for supplying per unit of water. In this case, energy consumption in this study means the energy consumed by the water supply sectors only. However, in line 319, the authors introduce the energy consumption by the steel and petrochemical sectors. I think more clarifications are needed. In addition, would the situation of energy shortage have a negative effect on water supply? There is no energy considered in equations 2-7.
- Equation 11. Similar to comment (3), technology development is supposed to benefit crop yield, but the exponential term here is dampening the crop yield.
- Environmental awareness put forward by van Emmerik et al. is intended to capture human sensitivity to environmental deterioration. In this study, the authors quantify environmental awareness by water shortage, food shortage and energy shortage (i.e., equation 14). I feel food shortage and energy shortage are more like social problems rather than environmental problems. It might be better if the authors change a name for this variable.
- Equation 18, 19 and 20 should be piecewise equations. I.e., when E is smaller than Ecrit, f(E) should be zero.
- Equation 21-23. If GDP would have an effect on water, food and energy systems, I think it might be more reasonable to use the magnitude of GDP instead of its changing rate.
- Section 3. Human response to the issues of water, food and energy shortages is an important aspect of the model. I suggest the authors give some observable evidences to show human adaptive response towards the mismatch between demand for and availability of water resources. for example any policy?
- A more detailed description of figure 3 is needed.
- Table 2. These are parameters and they may need to be listed in table 3. In table 2, the authors may need to show the initial conditions of the state variables, i.e., population, GDP, crop area, etc.
- Table 2 and 3 are too simple. At least the authors need to give a brief description of these parameters, as it is in table 5.
- There are only ten years data (i.e., 2010-2019, in yearly time step), but there are 35 parameters that need to be calibrated, which means this is a very complicated overparameterized model. I guess most of the parameters are insensitive. Perhaps an initial sensitivity analysis is needed to screen out those insensitive parameters before conducting calibration.
- Section 4.3. The authors explore the system sensitivity to seven parameters. I wonder why these seven parameters are chosen? Especially, all of them are threshold parameters. Are there any management implications obtained? I think it might be more informative if the sensitivities of the parameters related to human management actions are explored.
- Table 6. I am a bit confused about how the shortage rate is calculated. In some cases, the shortage rate is derived by dividing shortage by demand, and in some cases it is not. For example, in scenario I, the shortage of rural users is 0, why the shortage rate is 0.23%?
Additional minor comments:
- Line 63. The authors claim that system of systems model and agent-based model do not consider the feedbacks of integrated systems. I do not think this is true. A more appropriate literature review may be needed.
- In equation 4, crop area is denoted by A, but in equation 12, it is denoted by CA. please make it consistent.
- Line 251. The authors claim that environmental awareness proposed by van Emmerik et al. is more specific than community sensitivity. This is not the case. In fact, community sensitivity is proposed by Elshafei et al. through a more extensive literature review, and it is considered more sophisticated and is used more widely.
- Figure 4. Please try not to use abbreviations in the figure. It is very difficult to read.
- I notice that in some places, the authors use the word “resilience”. This is a complex concept, and as it is not the focus of this study, I suggest the authors use some simpler words.
- AC1: 'Reply on RC1', Yujie Zeng, 05 Dec 2021
-
RC2: 'Comment on hess-2021-521', Anonymous Referee #2, 18 Nov 2021
This manuscript presents a new approach for modeling water-energy-food nexus by incorporating social feedback loops driven by environmental awareness and a water resources allocation model into the system. It’s a interesting topic for researchers in the related areas, and the proposed approach has potential application value in other basins. The manuscript is clearly organized and the study background is described comprehensively in the Introduction. However, the method is not clearly explained in some places, and there are some detailed errors in words. Below are some detailed comments:
- The impact of water supply on energy consumption is related to industrial water, not ecological water or domestic water. Please clearly distinguish the impacts of different types of water supply on energy and food.
- In Figure 1, is the output of the water resources allocation model a total water supply or water supply of different sectors for every operational zone?
- In the energy system module, water supply not only affects energy consumption, but also energy supply, such as in thermal power, hydro-power and some other sectors. It is need to consider the impact of water supply on planning energy production.
- Please explain why GDP will affect the change of water quota in detail and provide some references for it.
- Line 197-202: There are several variables in the equation (6) that are not explained.
- For equation (9), why does the energy use quota of an optional zone multiplied by the water use quota of an optional zone equal total energy consumption? What is the definition of energy use quota in the paper? Please explain it.
- Line 238: the calculation formula of WSR isn’t presented in the paper, please add it.
- Line 328-331: Please add references to illustrate the contradictions between the increasing demands and limited resource supply will be aggravated in the study area.
- Are the impact of policy on water supply taken into account in the water allocation model, such as total quantity control of water consumed in the region?
- Line 358: How long is the data used for parameter calibration? Please add it.
- The conclusion section is too long now, please make it conciser and highlight the key conclusions.
technical comments:
- Line 124-125: There is no need to use the serial numbers "(1), (2)..." here, please getting rid of them.
- Line 174: The sentence “...are the of population...”is lack of some words.
- Figure 1: “Municipal water demand” projected by population is lack of rule water demand, which needs to be added.
- The font size of Equation (3)is not consistent with other equation
- Figure 4(i) : The text after “phase 1: ”is blank.
- Line 404: The word “phase”doesn't need an s.
- AC2: 'Reply on RC2', Yujie Zeng, 05 Dec 2021
-
RC3: 'Comment on hess-2021-521', Anonymous Referee #3, 20 Nov 2021
The authors create a multi-sector system dynamics model, including environmental awareness dynamics and coupled reservoir simulation. The model simulates, among other things, water demand, energy consumption, food production, environmental awareness, and population and GDP growth. The authors apply their model to the Hanjiang river basin and discuss the model simulation results at length. They identify stages of expansion, contraction, recession, and recovery for future water and energy dynamics as well as stages of expansion and stabilization for future food dynamics. The authors conduct a one-at-a-time parameter sensitivity analysis and also show that WEFS (water-energy-food-society) outcomes are strongly impacted by the presence or absence of reservoirs.
While this work aims to contribute in two primary areas – improved understanding of the impact of (1) environmental awareness feedbacks and (2) water supply reservoirs on WEF systems – I believe the work does not achieve these contributions, for the reasons described below:
- It is not clear what exactly about the approach is new. What separates the present study from those WEF studies cited in the introduction, other than the specific context and states modelled? It seems to me that the intended novelty might be coupling a WEF “system-dynamics” model with a detailed reservoir network simulation model, though this is not made clear in the paper. The discussions of model formulation and results do little to emphasize reservoir impacts, though the title suggests that reservoir impacts are central to the paper.
- Socioeconomic model (section 2.1.1, equations (2)-(5)):
- The model formulation and justification overlooks well-established growth models subject to resource constraints. Why not use a logistic model for growth?
- Each of these growth rates seem likely to be as or more effected by the *actual* resource limitations (i.e. shortages) than by the “environmental awareness” of those limitations. Yet, the physical limitations are not factored into these equations.
- I believe rates of change should be proportional to the state at time t, not the initial condition.
- The impact of technology development is either formulated unrealistically or discussed inaccurately – current formulation/discussion implies that technology suppresses growth.
- The water quota dynamics are especially unjustified – an exponential growth/decay model seems ill-fit.
- Water shortage model (section 2.1.2, equations (6)-(7)):
- The index for summation is not declared, making the equations difficult to interpret.
- The variable definitions are inconsistent and contradictory. Wdem is said to be water demand in line 201, yet WD also appears in equation (7) and is defined as water demand. There is also a Wd variable which is never defined.
- The temporal resolutions (time step and sub time step) are not explained and are therefore confusing.
- The distinction between “natural” and “total” water inflow is unclear.
- Energy system and Food system modules (sections 2.3 and 2.3, equations (8)-(13));
- These modules apply opposite approaches, without justification. The energy module simulates energy demand and takes energy production as an input (“planning energy production”). In contrast, the food module simulates food production and takes food demand as an input (misleadingly named “planning food production”). Why not simulated food demand or energy production?
- No justification is provided for formulating energy demand as a function of water supply, as opposed to population or GDP for instance. Water supply seems like a more important factor for energy production, though energy production is not modelled.
- I would think that the entire crop yield dynamics are due to technology changes (ignoring water shortage), yet technology change is offered as a single term in equation (11).
- From the results (Section 4, see especially Tables 2 and 5), it seems that a constant energy production and constant food demand are used to drive the model simulation. This seems unrealistic.
- Model validation (Section 4.1):
- The methods used to develop the observed time series are unclear. For instance, how exactly were the agricultural water demand exceedance frequencies used?
- The observed data is not sufficient to validate the model. The observed data cover a short period during the beginning of the simulation during which all states increase approximately linearly. The effects of shortage and environmental awareness are minimal during this period (as stated by the authors in their interpretations); therefore, the observations offer no validation of the awareness dynamics or feedback. That the model matches observed dynamics under this narrow, early set of conditions does not mean that the model can reliably simulate dynamics under drastically different conditions. For instance, a model which predicts perpetual linear growth in all states would seem to match the observations equally well. Given that the data does not validate the model, the model results are only useful to the extent that the model formulation seems true-to-reality. However, little justification is given for the model formulation, and as described above, there are many problematic elements of the model formulation.
- Model results (Sections 4.2-4.3):
- Most of the discussion of the results (co-evolution of WEF system) is a text description of what is seen in the figures. The discussion does little to draw out and emphasize insights.
- The sensitivity discussion does little to add understanding. Most interpretations of sensitivity results are vague, such that the same observations could be stated just from the variable definition and model formulation. For example, in lines 551-553, the effect of lowering the food shortage sensitivity threshold level is obvious from its definition.
- Impacts of reservoir system (section 4.4):
- The methodology applied here is unclear, what exactly does it mean that one scenario considers allocation and the other doesn’t?
- Nonetheless, it seems that scenario I is running the model with the real-world reservoir network and scenario II is running the model with all reservoirs removed (?). If so, scenario II does not seem like a useful comparison. Is the region considering removing any or all dams in the basin?
- There are language issues throughout the manuscript – most frequent were typos, poor sentence structure (lots of passive voice that creates confusion about who is the subject and what exactly they are doing), and inappropriate word choice. There are too many to list specifically.
- AC3: 'Reply on RC3', Yujie Zeng, 05 Dec 2021
-
RC4: 'Comment on hess-2021-521', Anonymous Referee #4, 20 Nov 2021
This study modeled the WEF nexus by incorporating community sensitivity and reservoirs operation, where the co-evolution behaviors of the nexus across the water, energy, food, and society (WEFS) were simulated by the system dynamic model. The proposed approach was applied to the mid-lower reaches of the Hanjiang river basin in China, and the results indicated that water resources allocation could ensure water supply through reservoirs operation and greatly decrease the water shortage rate. This study is an interesting and crucial one for improving resources management. While modeling the WEF nexus in a large watershed is a very challenging problem and difficult to validate its suitability and applicability, especially when there are only limited datasets. This study made a great effort in this direction and proposed a sophisticated methodology with some preliminary analyzed results, which is a valuable contribution to the scientific community. However, I have some concerns and suggestions, which need to be better addressed, listed as follows.
- The initial conditions of external variables for the integrated system shown in Table 2 and the calibrated parameters presented in Table 3 should be explained in more details. I am curious why many parameters have the same calibrated value. How to set these values?
- How many datasets are used for model calibration? The number of calibrated parameters used for model calibration should be discussed. How to justify the suitability and applicability of the calibrated model should be given.
- The “Respond links” among the different variables in the WEFS nexus should be explained in much more detail. The terms of feedback functions based on previous work should further explain their suitability.
- Figure 4 shows the trajectories of population, GDP, crop area, water demand, energy consumption, food production, shortage rates for water, energy, and food, awareness for water shortage, energy shortage, and food shortage as well as environmental awareness during 2010-2070. The trajectories are the basis of the following analyses. How to get these trajectories should be given in more detail, and their suitability should be discussed?
- How to divide the evolution of water demand and energy consumption into four phases should be given?
- The seven controllable parameters adopted for sensitivity analysis should be discussed in more detail.
- The conclusion seems like a long summary of the current study. The main contribution with brief (solid) scientific findings extracted from this study might be more interesting to read and easy to learn.
- AC4: 'Reply on RC4', Yujie Zeng, 05 Dec 2021
-
EC1: 'Editor Comment on hess-2021-521', Murugesu Sivapalan, 05 Dec 2021
The paper has benefited from excellent and constructive comments from 4 reviewers. The comments are not only favorable but also constructive, with Reviewer #3 offering some serious criticisms. The authors have responded well to the concerns raised and have indicated a serious attempt to address these. I welcome the offer to substantially revise the paper, including some additional work. I look forward to the revised version, which I will ask at least two of the original reviewers to critically evaluate again, before making the decision to publish in HESS. Thank you
-
AC5: 'Reply on EC1', Yujie Zeng, 18 Dec 2021
Dear professor Murugesu Sivapalan:
Thank you very much for giving us the opportunity to submit our revised version. We have carefully revised our manuscript according to the constructive comments from 4 reviewers. Please find our revised version and responses in the attached file.
Sincerely,
Yujie Zeng
-
AC5: 'Reply on EC1', Yujie Zeng, 18 Dec 2021
Yujie Zeng et al.
Yujie Zeng et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 880 | 226 | 20 | 1,126 | 7 | 8 |
- HTML: 880
- PDF: 226
- XML: 20
- Total: 1,126
- BibTeX: 7
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1