Assessment of consistency in water allocation decision making

The case of the Murrumbidgee Regulated River in Australia


Tesis de Máster, 2018

115 Páginas, Calificación: 9.00/10.00


Extracto


Table of Contents

Abstract

Acknowledgements

List of Figures

List of Tables

Abbreviations

Introduction
1.1 Background
1.2 Problem Statement
1.3 Motivation of this Study
1.3.1. Expected societal benefit from the research
1.3.2. Expected scientific benefit from the research
1.4 Research Objectives and Question
1.5 Report Structure

Literature Review
2.1 Water allocation
2.2 Water allocation process in different region
2.3 Use of additional hydrological Information:

Case Study Area
3.1 The Murrumbidgee River
3.1.1. Catchment description
3.1.2. Origin and flow direction
3.1.3. Regulated Murrumbidgee River
3.1.4. Burrunjuck Dam
3.1.5. Blowering Dam
3.1.6. Climate
3.1.7. Land Use
3.1.8. Irrigated agriculture
3.1.9. Murrumbidgee Irrigation Area
3.1.10. Coleambally Irrigation Area
3.1.11. Irrigation demand
3.2 Stream flows

Water Allocation in the Murrumbidgee
4.1 Water Sharing
4.1.1. Basic water rights and license
4.1.2. Entitlement of License categories and allocation
4.1.3. Available water determination (AWD)
4.1.4. Revision of the AWD
4.1.5. Supply of allocated water
4.2 Water allocation rules
4.2.1. Environmental Water
4.2.2. Allocation to Water Access License (WAL) categories
4.2.3. Carryover
4.3 Water allocation process

Methodology
5.1 Research Design
5.1.1. Activity Diagram
5.2 Water sharing rules and allocation process
5.3 Data Collection
5.4 Water allocation model
5.4.1. Input data
5.4.2. Model setup:
5.4.3. Resource calculation
5.4.4. Available water determination for license categories (AWD)
5.4.5. Model performance evaluation
5.5 Consistency in water allocation decision making
5.5.1. Framework for consistency assessment
5.5.2. Hypothesis regarding best and worst scenario
5.5.3. Step-1: Representing allocation decision in a step diagram
5.5.4. Step-2: Measurement of area under the curves
5.5.5. Step-3 : Measurement of Consistency
5.5.6. Hypothesis judgement
5.6 Application of additional hydrological information in water allocation

Results
6.1 Available Water Determination (AWD)(2012-13)
6.1.1. Storage
6.1.2. Expected Inflow
6.1.3. Reserve and Losses
6.1.4. Allocation to the general Security
6.2 Comparison of Initial AWD in different years
6.3 Revision of the allocation decision:
6.4 Historical water allocation to GS
6.4.1. Water allocation to GS in very dry years
6.4.2. Water allocation to GS in dry years
6.4.3. Water allocation to GS in wet years
6.4.4. Summary of the analysis
6.5 Emulation of water allocation process to GS
6.6 Allocation and Release
6.7 Consistency in GS allocation in previous years
6.8 Model simulation with different inflow data
6.8.1. Simulation result using perfect information
6.8.2. Simulation result using conservative inflow
6.8.3. Simulation result using inflow data of W3RA global model
6.8.4. Simulation result using inflow data of HTESSEL global model
6.8.5. Simulation result using inflow data of SURFEX-TRIP global model
6.8.6. Simulation result using inflow data of PCR-GLOBWB global model
6.8.7. Simulation result using mean inflow of global model datasets (Mean-GHM)
6.9 Relative consistency index (RCI)

Discussion and Conclusion
7.1 Discussion
7.2 Conclusion
7.3 Limitations and assumptions of the study
7.4 Recommendation for future work

Reference

Appendices
Appendices A Water Allocation Data (2014-15)
Appendices B Example of a Water Access License in NSW
Appendices C Longitudinal Profile of the Murrumbidgee
Appendices D Model Input data

Abstract

Decision making in water allocation process to irrigated agriculture is a complex job due to the increasing demand of competing sectors and the uneven distribution of usable water resources both in time and in space. One of the most important parts of this job is the quantification of the available water from all sources to meet the demand to the end of the irrigation season. In doing so, decision makers are often very conservative to increase the reliability of allocation process and decrease the probability of damage associated with failure to deliver the allocated water.

The water allocation process in the Murrumbidgee region, New South Wales, Australia is well monitored, and follows an established set of clear rules and supported by various hydro- meteorological data. Despite this, the water allocated to irrigated agriculture, in particular to annual crops, is often conservatively low at the beginning of the season. This allocation will then be increased, as a results of the revision of allocation volumes all through the cropping season. As a consequence of these inconsistencies between the initial and the final allocation of water, the optimal benefit from the available water is hampered.

This research reviews the water allocation process in the Murrumbidgee region, identify the reason behind the inconsistency, develop a framework to measure the consistency and investigate the usefulness of additional hydrological information to improve the consistency of the allocation process. The water allocation process from 2011 to 2016 has been reviewed, and based on this understanding and available data, a water allocation model has been developed which is able to simulate the allocation process for the region. The model performs well in emulating the previous allocation process and can be tested with any datasets. Moreover, a framework has been developed to measure the consistency of allocation decision which is expressed in a consistency index.

Results show that the consistency in allocation process varies depending on the inflow condition into the catchment. Water stored in two upstream reservoirs and predicted inflows for the rest of the years into those reservoirs are the main source of water for allocation in a given year. We found that the predicted annual natural inflow into the reservoir are currently taken as the observed historical minimum inflows. However, in most years, the observed inflow exceeds this expected inflow, which provides the additional water available for allocation as the year progresses. As a consequences, allocation volume to the least priority user, agriculture, also increases as the year progresses.

Based on the allocation model that we developed, we investigate whether using inflow data simulated by global model as additional information can improve the consistency of the water allocation process. We used simulated runoff from three land surface models; HTESSEL, SURFEX, W3RA, and one global hydrological model, PCR-GLOBWB, to estimate the total inflow in the catchments of the two upstream reservoirs. These estimates are then used to simulate the allocation process for three consecutive year, starting from 2011-12. We found that by using this additional information, a better consistency in the allocation decision could be achieved. It implies that the information regarding end of season allocation can be achieved at the start of the cropping season and allocation volume remain almost stable from the start to end of the year. It can assists farmers to make cropping decisions confidently at the very beginning of the year, and thus can contributes to enhance agricultural production as well as ensure optimal benefit from the available water resources.

Key Words : Water allocation, Murrumbidgee, General security, Murray Darling, consistency, decision making, Burrinjuck, Blowering, rice production, availability of water, New South Wales, NSW, Australia.

Acknowledgements

I would like to express my gratefulness to my mentors Alex Kaune Schmidt, Dr. Micha Werner and my supervisor Prof. Charlotte de Fraiture. Their advice and guidance throughout this research time not only make it possible to come into this ending point but also make me more competent to face future difficulties to manage water related issues in my professional life. Especially, I will remember few words of them “…at the end of the day, you have to solve your problem…”, “…be critical in your work…” and “…try to write your understanding in your own words…”.

I am grateful to my elder brother, Dr. Mohammad Russeel Chowdhury, who has managed our family in absence of me and always pushed me forward throughout my study life. I also express my gratitude to my beloved mother, wife and son for waiting so long and wishing my safe return in every second. I also pray for peaceful afterlife of my dearest late Father who decided to put me in a good school thirty years back.

In this period of thesis work, I read many reports, scientific publications and forum discussion. Though I have mentioned some of those in the reference list, but most of those will be remain alive in my understanding of water related issues. I would like to acknowledge the financial support of Joint Japan/World Bank Graduate Scholarship Program and institutional support of UNESCO-IHE. I would also like to thank Bangladesh Water Development Board authority for continuing my salary and position in the department and thus providing scope to learn advance technologies for becoming more informed to work for the betterment of Bangladesh.

Last, but not least, I would like to thank all of my friends in UNESCO-IHE, specially Mr. Jamal Haider, Mr. Feroz Islam, Mr. Hafiz Ijaz Ahmed, and Mr. Junaid Ahmad for their cooperation and memorable company in the Netherlands.

Mohammad Faysal Chowdhury

Executive Engineer (Civil)

Bangladesh Water Development Board. Delft/ The Netherlands, 26th March, 2018.

List of Figures

Figure 1-1: Improvement in Water allocation.

Figure 1-2: Revision in water allocation to general security (2010-11)

Figure 1-3: Comparision between initial and final GS allocation volume from the year 2004 to 2016. . 4

Figure 1-4 Allocation volume at different time of the years and annual usage .

Figure 1-5: Relation between rice area harvested and GS allocation

Figure 3-1: Location of the Murrumbidgee valley .

Figure 3-2: Topography of the Murrumbidgee catchment .

Figure 3-3: The Murrumbidgee River .

Figure 3-4: Major dams and weirs in the regulated Murrumbidgee

Figure 3-5: Average annual rainfall in The Murrumbidgee Catchment

Figure 3-6: Average monthly rainfall at Tumbarumba, Wagga Wagga and Hay.

Figure 3-7: Annual evaporation in the Murrumbidgee catchment .

Figure 3-8: Land use pattern in Murrumbidgee region .

Figure 3-9 Major Irrigation area in the Murrumbidgee region.

Figure 3-10: Irrigation demand in two major irrigation area (2000-01).

Figure 3-11 : Long term mean monthly river flow at different location (GL/month) .

Figure 3-12 : Daily and annual discharge at Wagga Wagga from 1899 to 2018.

Figure 3-13 Mean daily flow at different location along the Murrumbidgee river. .

Figure 4-1: Condition to extract water by Domestic and stock right.

Figure 4-2: Entitlement and allocation for Water Access Licenses (WALs).

Figure 4-3: Computation of available water for consumptive use in the Murrumbidgee regulated river

Figure 4-4: Priority and entitlement of license categories .

Figure 4-5: Water allocation simplified.

Figure 4-6: Carryover rules for General Security and conveyance license.

Figure 4-7: Annual accessible water calculation process.

Figure 4-8: Available water determination (AWD).

Figure 5-1: Sequence of activities followed in the research

Figure 5-2: Blowering daily storage and outflow (July 2007 to July 2017).

Figure 5-3: Burrinjuck daily storage and outflow (July 2007 to July 2017).

Figure 5-4: Model workflow .

Figure 5-5: Reservior operation in each time step .

Figure 5-6: Average inflow and demand curve used in the model setup.

Figure 5-7: Cumulative volume of water for allocation in the model.

Figure 5-8: Available water determination for license categories in the model.

Figure 5-9: Measurement of consistency for two allocation scenario .

Figure 5-10: Hypothesis on consistency measurement .

Figure 5-11: Preparation of Allocation Curve.

Figure 5-12: Preparation of area under curves.

Figure 5-13: Catchment shapefile of the Blowering and Burrinjuck reservior.

Figure 6-1: Determination of water that will be available in the year 2012-13. .

Figure 6-2 Starting (1stJuly, 2012) AWD to various categories of WAL

Figure 6-3 Initial AWD for the year 2010-11 to 2016-17.

Figure 6-4: Comparison between observed and expected inflow.

Figure 6-5: Comparison of actual natural inflows with average inflow.

Figure 6-6: Total annual natural inflows into the Murrumbidgee valley.

Figure 6-7: Historical water allocation for GS in very dry years.

Figure 6-8: Historical water allocation decision for GS in dry years.

Figure 6-9: Historical water allocation decision for GS in Wet years.

Figure 6-10: Actual observed and modelled GS allocation (2011-2016).

Figure 6-11: Unused water and Carryover .

Figure 6-12: Reservoir storage for different release-allocation ratio.

Figure 6-13: Simulation result using perfect inflow data (2011-14)

Figure 6-14: Simulation result using conservative inflow data (2011-14)

Figure 6-15: Simulation result using inflow data of W3RA model (2011-14) .

Figure 6-16:Simulation result using inflow data of HTESSEL model(2011-14) .

Figure 6-17: Simulation result using inflow data of SURFEX-TRIP model

Figure 6-18: Simulation result using inflow data of PCR-GLOBWB model (2011-14) .

Figure 6-19: Simulation result using mean inflow of global model datasets (Mean-GHM) .

List of Tables

Table 4-1 Total Entitlement for Access Licenses in the Murrumbidgee Regulated River .

Table 5-1: Details of data acquisition station .

Table 5-2: Overview of Global Models .

Table 5-3: Inflow data for different simulation scenario.

Table 6-1: Consistency of GS allocation decision making (2004 to 2016)

Table 6-2: RCI value for different global inflow data: .

Abbreviations

Abbildung in dieser Leseprobe nicht enthalten

Introduction

This chapter provides the background of the research. It consists of the problem description that motivates the research, the objective of this research and the questions addressed through this research.

1.1 Background

Despite the fact that almost two thirds of the land surface of planet earth is covered by water, we still recognize it as a scarce resource. On the other hand, though experts have already proven that not a single droplet of water can vanish from this earth by doing mass balance calculation, people are experiencing increasing shortage of water day by day all over the world. Common answer to this challenge is too little a percentage of readily available fresh water with respect to total available water and increasing demand for usable water. Though it is difficult to increase the percentage of fresh water, the feeling of shortage can be reduced by managing the available water wisely.

Water allocation is a fundamental part of water resources management (Patrick, 2012). To balance between demand and supply, decision makers often categorize the users based on their importance. Priority is often fixed by consultation with the stakeholders. For instance, if a water source has to meet the demand for a city as well as for agriculture and industrial uses, most the time high priority is given to the city water supply, daily drinking necessity or domestic uses. This priority based water allocation is commonly practiced in water scarce area and well recognized all over the world. The process of water allocation is highly dependent on the availability of water. Decision makers have to assess the availability of water with adequate information to enhance the reliability of water allocation system and ensure optimal allocation to competing demands.

In many areas, available water is distributed among the users with daily necessities attended to first, and the remaining volume allocated to seasonal users such as agriculture. In agriculture, demand varies a lot due to the inequality of the water requirement through the growing period of crops, and the allocation process considered difficult as a consequences. Assessing available water as well as the future inflow into the system is vital part of this tricky job. If water allocated to a crop cannot make available at the right time, the result will be severe crop damage. On the other hand too little allocation during the planting season and too much availability of water during the rest of the season; causes suboptimal use and loss of benefit.

This thesis focuses on the analysis of the decision making process in water allocation in a major river basin in Australia, the Murrumbidgee. This includes the development of a frame work to measure the inconsistencies or variation in the water allocation process and evaluating inter annual allocation decisions. Moreover, the thesis explores how additional hydrological information can affect the decision making process.

1.2 Problem Statement

The Murrumbidgee River in NSW, Australia carries water from upstream, where it is regulated by two major dams and acts as the main source of water for irrigation, local water supply, industrial demand, daily needs and ecosystems in the region. To extract water from the river, users need a water access license issued by the government of NSW. These licenses are associated with certain number of unit shares, or a water volume entitlement (maximum 1ML/year for each unit share) that defines the maximum amount of water that can be extracted from the river in a year. However, the actual amount of water that will be available to extract in a certain year depends on the allocation volume to licensed users, which varies from 0 to 100 percent of the entitlement based on the availability of water. In this region, water is allocated based on the priority of the user’s water access license category, as defined in the water act and water sharing plan. For irrigators, there are two types of license known as high security and general security. High security licenses are mostly for irrigators with permanent crops (i.e. vineyards, citrus, orchards etc.) where the investment is very high and that require a constant annual application of water to remain productive. General security licenses are for growing annul crops such as rice, wheat, corn, tomatoes, cotton etc. As per government rule, high security access license category get higher priority than the general security in case of water allocation. It means that after allocating required volume of water as mentioned in the access license of high security category, remaining available water is given to the general security license holder. Hence, in case of shortage of water than the total requirement, general security get less amount than the entitlement.

In any water year, which spans from July to the following June, irrigators get the information about the quantity of water they will get in that year by an allocation announcement issued by the NSW government. At the beginning of each water year on July 1st, an initial allocation announcement is made based on the availability of water in the valley. Water availability is mainly calculated on the basis of stored water in several storages and expected future inflow into the major storages throughout the year. While calculating future inflow into the storages, the lowest historical inflow observed in the last 120 years is taken into consideration (DPI Water, 2013). The main reason behind this conservative manner is to ensure delivery of the initial announced water confidently to the users over the year and reduce the possibility to decrease the initial allocation volume from the user accounts, unless the inflow sequence drops below the historical minimum. As rainfall in the area varies a lot from year to year as well as resulting inflow into the storages, the minimum inflow volume assumption makes the initial allocation more reliable. However, this conservative approach provides lower volume of water than the total issued entitlement. As a consequence, after allocating water to high security category, general security license holders get less amount of water than their entitlement.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1-1: Improvement in Water allocation.

( A t the start of the year, initial allocation is low but as soon as availability of water increases, allocation also increases)

In reality, most of the time inflow into the storages typically exceeds the minimum historical inflow sequence and more water becomes available as the season advances. Under this improved condition, general security license holder get an additional amount of water added with the previous allocation volume. The authority assesses the inflow of water into the storages regularly and whenever observed inflow exceed the expected inflow, allocation to the general security increases accordingly. Thus, increase of allocation to general security continues based on the inflow condition until the cropping season (usually February) ends. Because of this revision process, significant differences between the initial allocation and end of season allocation have been observed in many years. An illustration is shown in Figure 1-1, a general security license holder has 100 ML as entitlement but authority allocate 50 ML during initial allocation. Afterwards, when the availability of water increases, allocation raised to 70 ML. Allocation to general security for the year 2014-15 is shown in Figure 1-2. The initial allocation to general security (GS) was 480 GL which had been revised several times and increased by 1,416 GL within few months as the season progressed and finally reached as high as 1,896 GL by mid of December. Likewise, almost every year, there is significant increase in GS allocation from the initial announcement as can be seen in Figure 1-3.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1-2: Revision in water allocation to general security (2010-11).

( In 2010, the initial allocation to General Security was 480 GL which

has been increased by 1,416 GL within few months as the season progressed and reached as high as 1,896 GL) (data source : water.nsw)

Abbildung in dieser Leseprobe nicht enthalten

Figure 1-3: Comparision between initial and final GS allocation volume from the year 2004 to 2016.

( Every year, final GS allocation volume is more than the initialalocation volume) (data source : water.nsw)

Usually, farmers start to prepare the land for cultivation at the mid of July and sow summer crops by October. Based on the individual type of crop’s water requirement, they have to estimate the quantity of planting with respect to the amount of water that will be available till the end of season. If they do the planting according to the early water allocation announcement then the increase in water allocation will cause over irrigation and become useless. On the other hand, if they plant crops based on expectation of a certain amount of additional allocation as the season develops but the actual allocation volume do not reach that level, there is a high risk of crop damage and associated financial loss.

1.3 Motivation of this Study

Irrigated crops contribute a lot to the economy of the Murrumbidgee region. Regional processing industries are also highly dependent on the produced crops like rice and horticulture. Hence, it is obvious that regional income and employment has a strong correlation with the agricultural production.

Abbildung in dieser Leseprobe nicht enthalten

Figure 1-4 Allocation volume at different time of the years and annual usage

[ ( A)Total usage from the start of year till the end (blue line) is much lower than the end of season allocation (yellow line) but closer to the early season allocation (left figure). (B) Co-relation between allocation volume up to 15th august and total water usage is highest (88%) (data source : water.nsw)]

In the Murrumbidgee region, it is observed that total annual usage of water under GS is highly dependent on the July-August allocation volume. As shown in Figure 1-4, total annul usage of GS is always lower than final allocation volume. The correlation value between average usage and allocation increases up to mid of august (from 82% to 88%) and afterwards the value started to decrease (from 87% to 79%). It implies that farmers usually take planting decision based on the early season allocation to reduce the risk of cropping failure. As a result, the volume of water that is allocated afterwards remain unused and production does not reach to its maximum level. Moreover, due to lower percentage of water allocation in early season, farmers find it more profitable to sell their limited water allocation to other users like horticulturalists (Rice Industry, 2010). However, if the farmers would have adequate information at the starting of irrigation season about the total water that they are going to have by the end of cropping season, they may plant more and the resulting crop production would go up. This will ensure the optimal use of limited water resources and greatly benefit the regional economy. Therefore, to investigate whether better knowledge on end of season allocation to GS can be achieved by using additional hydrological information at the beginning of the year, research is felt to be necessary.

It is also found that sometimes the initial allocation volumes are comparatively higher and subsequent increases are lower. As a result, the number of revision as well as the difference between early and end of season allocation becomes smaller. On the contrary, in some years the initial allocation is lower but subsequent increases are higher, resulting in a large difference between early and final allocation as well as a large number of revisions. This calls for a framework to be developed to evaluate each year’s allocation process as well as to compare between different year’s decision making processes. However, there is currently no measuring tool to identify how much variation in water allocation decision has occurred over the last years. Hence, decisions makers cannot evaluate the inconsistency that has been going on in the water allocation process over the years.

The lack of scientific research concerning the issues described above is the motivation for the current research.

1.3.1.Expected societal benefit from the research

Abbildung in dieser Leseprobe nicht enthalten

Figure 1-5: Relation between rice area harvested and GS allocation

(Rice cutivation increase with the increase of water allocation to GS) (Data source agriculture.gov.au)

Murrumbidgee valley is very productive for rice cultivation. Rice produced in this valley not only meet the local demand but also contributes in economy by exporting to other countries. Almost 1.3 ML of water is required to produce 1 ton of rice in Murrumbidgee valley (DPI, 2017). As shown in Figure 1-5, rice cultivation area increases with the increase of GS water allocation in the valley. It implies that, If the farmers get end of season allocation information at the start of season, they can produce more rice as well as other crops. It will enhance the economy of the farmers and boost the agricultural dependent industry in Murrumbidgee. As a consequence, enhancement in social and food security can be achieved.

1.3.2.Expected scientific benefit from the research

Water allocation process in the Murrumbidgee valley is well monitored with a set of well- established rules that is enhancing the economy and protecting the environment at the same time with the limited water resources. Lots of region in the world is experiencing problem in water allocation that can be overcome if the process is well organized and monitored properly. This research has explored the water allocation rules and process in the Murrumbidgee which can be useful for the research community to apply this principals in other relevant region for maximizing benefit with limited water resources. The allocation model that has been developed in this research can be tested with any data set to investigate the annual allocation outlook at the very beginning of the year. Moreover, the development and working principle of the allocation model can of great use to develop similar type of model for other region. In addition, usefulness of additional hydrological information freely available from global hydrological models has been also assessed in this research to improve the water allocation process in the Murrumbidgee. It signifies the benefit of global datasets which will give evidence to the researcher to use this data in other area.

1.4 Research Objectives and Question

The overall objective of this research is to evaluate the consistency of the water allocation process in a regulated river like the Murrumbidgee and assess whether this can be improved by additional hydrological information. The specific objective and the research questions that this research aims to answer includes:

- To review the water allocation rules in the Murrumbidgee regulated river and evaluate how water is allocated to the different users throughout the (water) year and how much it varies from the initial water allocation announcement.

Research Question 1-: How can the water allocation decisions that were made based on historical data in the regulated Murrumbidgee be evaluated?

- To develop a framework to measure the consistency of water allocation decisions.

Research Question 2-: How can the consistency in the decision making process be assessed in a regulated river like the Murrumbidgee?

- To assess how additional hydrological information can contribute to support water allocation decisions.

Research Question 3-: Can additional hydrological information contribute to more consistent water allocation decisions?

1.5 Report Structure

Chapter 2: Literature Review

This chapter provides a description about the components and various aspect that are considered for improving the water allocation process. In addition, it also briefly describes various mechanism of water allocation that has been practiced in different regions. The significance of global model data is also mentioned.

Chapter 3: Case Study Area

This chapter aims to provide a brief description about the study area that includes catchment, river flow, climate and land use. Moreover, regulation along the river including two major reservoirs are highlighted. The existence of major irrigation districts and their demands are also explained.

Chapter 4: Water Allocation in the Murrumbidgee

This chapter provides a brief description about the water sharing process in the Murrumbidgee. It also describes various license category, their priority, entitlement and the rules followed in available water determination (AWD) to allocate water to different categories. Water sharing for environmental purposes and carry over rules are also explained. It comprises of an analysis of sequential process followed in water allocation with its components.

Chapter 5: Methodology

This chapter provides a description of research methodology and techniques used to achieve the research objectives and address the research questions. It comprises data collection activities; development and principals of allocation model; development of framework for measuring consistency of water allocation decision process and additional hydrological data to assess the goodness of their use in the allocation process.

Chapter 6: Results

This chapter consists of the results obtained in this research work in connection to the objectives set and in answering the research questions. It includes detailed calculation of water allocation for the year 2012-13, comparison between different year’s initial allocation and the reason behind the revision of General Security (GS) allocation. An analysis of water allocation to GS with respect to the historical inflow conditions is also provided. Moreover, the model results in emulating the historical allocation and their consistency is included. Assessment of additional information in the decision making process is also explained.

Chapter 7: Discussion and Conclusion

This chapter contains brief discussion regarding this research work, summarizes the research findings and its relevance to the objectives. It also includes the limitations of this research and prospect of further research.

CHAPTER 2

Literature Review

T his chapter provides a description about the components and various aspect that are considered for improving the water allocation process. In addition, it also briefly describes various mechanism of water allocation that has been practiced in different regions. The s i gnificance of global model data is also mentioned.

2.1 Water allocation

Water is life and efficient water resources management is the key for sustainable development (Morris, 2003). Demand for water in all over the world is growing due to pace in urban development, food production, population and economic growth (GWP, 2009). Conversely, uncertainty in availability of fresh water has been increasing day by day due to climate change and anthropogenic activities. As a consequence, around 900 million people are living in areas with shortage of water and 700 million are moving forward to experience same in a great pace (De Fraiture, 2010).

Water allocation process includes assessment of available water and sharing it between legitimate users or sectors spread over a basin, sub-basin or region. While assessing available water in a basin, surface, groundwater as well as interbrain transfers are often considered. Moreover, to make equitable allocation of water between users or sectors, water sharing plan has been formulated in various part of the world. Those plans often make provision of water right in the name of permits, access license, concessions or entitlements to the users that authorizes users to store, regulate or withdraw water from the source. However, process and objectives of water allocation has been evolved in different part of the world in different ways (Le Quesne, 2007).

In an area where fresh water supply is more than the demand, water allocation process is not felt necessary. However, allocation process become tricky when availability and natural distribution of water could not meet all the demand from various sectors like domestic usage, industry, energy supply, agriculture, recreation and environmental requirements. While allocating water among the competing demand, various factors like equity between users, economic and social benefits, environmental protection and reliability of supply are considered to make the process more effective and efficient (Hellegers, 2015). Equity describes the fairness of the distribution of limited resources among the different sectors or regions and lack of which creates tension between the users. Economic and social benefits associated with the economic efficiency that represents productivity of the water from different sectors. Reliability depends on the balancing process between demand and supply so that the system don’t collapse even in severe shortfall due to natural variability of supply. As water is the main driving force for environmental sustainability, allocation should consider its needs to maintain dependent ecosystem as well as services like ground water recharge, sediment transport and waste assimilation (Speed, 2013).

Water allocation planning can be of long-term, medium term and short term (Kaune, 2017). In long term planning phase, such as developing an large scale irrigation area or basin level allocation, all the sources of water and historical variability of available water in those sources, future demand and supply estimation is crucial. After considering these factors, long term share of available water is given to the regions, sectors or individuals. This is commonly known as entitlement or water right which usually include the timing, location, volume or priority of water extraction. On the other hand, in medium or short term planning phase, such as making annual or seasonal allocation, based on current status of available water in the region, actual volume of water available for extraction is usually granted to each entitlement.

Effective water allocation and management requires better assessment and understanding of present as well as future water availability to meet the demands of various sector spread over the basin during all season (UNESCAP, 2000). Adequate information about the various sources and quantity of supply from each source may lead to make more efficient water allocation decisions (Kaune, 2017). Therefore, assessing availability of water resources has been an issue of importance in most countries all over the world. Availability of water determines the quantity as well as temporal and spatial distribution of water resources in a region. The most common way of estimating quantity of water in a river basin is the historical observed long term average annual flow. To consider the seasonal variability in the annual flow, average monthly runoff usually considered. Where the long-term data is not available, appropriate hydrological model for surface and/or ground water are often used to generate long term average flow and predicting future runoff as well.

In most part of the world, large portion of total estimated annual water become available in wet season when demand (i.e agriculture) declined, and conversely supply becomes low in dry season when demand grows high. Therefore, to accumulate wet season’s water and use it efficiently, most of the water scarce region regulate the river and store water by various means (dam, barrage, weir). It ensures more security for water availability for different sectors (Grill, 2015). The portion of total available water that become utilizable largely depends on the total storage capacity of the region. However, if the objective of storage is to hydropower generation or flood storage, the amount of utilizable water become low due to continuous release requirement for electricity generation or maintaining dam to reduced level to capture flood. Therefore, even though a basin having adequate amount of runoff may still suffer from shortage of water due to seasonal variability of available water, poor storage capacity or storage with different objective. After determining the utilizable volume of water considering all this factors, total allocable water can be determined by deducting environmental and downstream flow needs, various type of losses or another requirements.

Demand of water is generally estimated based on the water rights in the form entitlement or license in a basin or region. When all the demand cannot be met by the total water available for allocation, water dependent sectors has to be categorized based on their importance in the catchment and high priority sectors get the maximum assurance (Dinar, 1997) of water supplies even in scarce condition. Prioritization often done based on regional objective or susceptibility of the sector. In most of the water scarce region, highest priority is given to the basic human needs (drinking water, domestic usage, sanitary purposes,) supply followed by industry, agriculture, hydropower or other specific demand (UN, 2000). However, decision in water allocation among different sectors often seem to be contentious due to the increasing demand and reduced supply (Ramsar, 2010).

Agriculture consume nearly 80% of total global water uses (Molden, 2007). Farming system in a region is mostly dependent on the hydrological condition of that area. Rain fed farming system is very common in countries like Sweden, Poland and Belgium where rainfall distribution confirm required amount of soil moisture all through the cropping season. In contrast, irrigation form surface water bodies and groundwater is very common for agriculture in arid and semi- arid areas like Australia, Spain and Turkey. In monsoon climate, people in some area store wet season rainfall in reservoirs and meet the demand of dry season. There are some areas where desalinated water and recycled wastewater are also used to meet cropping demand (OECD, 2010). As irrigated agriculture is mostly dependent on the human intervention to get continuous supply of water, it needs much attention in water allocation process.

2.2 Water allocation process in different region

Water allocation process can be different in different location based on demand, climate, available infrastructure to monitor, store and regulate water. Region where user’s water share are defined as fixed volume of water/year, every year the allocation is same. For instance, in the Colorado River Compact, average of previous ten year water is allocated to each share. The shortcoming of this approach is; when supply of water fall below the average, users experience difficulties to cope with usual demand.

In some region with high degree of river regulation, availability of water is determined based on storage in the reservoirs and projected water that will be come from snow or rainfall. For example, in the Lerma-Chapala system in Mexico, the stored water in the Lake Chapala is added with the projected inflow into the system by comparing the current rainfall-runoff patterns with the historical record (CONAGUA, 2012) .

In some regions, to make the supply of water more reliable, only the water currently stored in the reservoir is taken into consideration when allocating water. Hence, no future inflow is considered and expected losses are excluded that ensure the deliverability of all allocated water to the users. For instance, water allocation from Driekoppies and Maguga dams in the Incomati River Basin only use the dam storages at the start of the year. Even more conservative approach are found in some arid region where rainfall variability is too high. In those area, total stored water is divided into the various demand for several years (even 3 years) and allocation is done in a manner so that all demands can be met for the next years even if no inflow happens during the long period. This type of approach is also applicable in case of increasing reliability of water supply to the users.

In some area, expected rainfall is considered for the whole year to allocate water. In that case risk of failure to make available of allocated water to the user increases. In this approach factors like variability of rainfall for longer period, probability of the occurrence of expected inflow into the system and consequence of failure of supple has to be considered. This type of aggressive approach is applicable for irrigating annual crops (i.e rice, wheat, vegetables) when farmers can take the advantage of higher production during wet years to compensate lower income during drought years due to low or zero allocation. However, this mechanism is not good for irrigating permanent crops (i.e horticulture, grapes, citrus) as periodic shortage of water may cause greater loss to the farmers that may need longer duration to overcome. Considering this factor, permanent plantings are allocated with lower volume with higher reliability form the available water to reduce the consequence of the failure of supply (Speed, 2013). For instance, in the Central Valley Project (CVP), California, initial allocation to various sectors at the start of each year is done based on a conservative estimation of water availability which includes reservoirs storage, snowpack and precipitation conditions, compared to the historical average in the Sierra Nevada and Central Valley. Allocation volume is usually re- examined on a monthly basis to revise the allocation considering various factors like reservoir storage, water right priority and overall hydrological condition (USBR, 2018).

In Oregon, USA users need to get permit or water right from the government authority to withdraw water from surface or groundwater sources. The amount of water that can be withdrawn is mentioned in the right and the oldest issued right (senior) get higher priority in withdrawing water. In case of users of surface water, authority issues total water right based on 80% exceedance flow in the river which means there is 20% chance for junior right holders not to get adequate amount of water in a year. Moreover, during drought condition, after meeting all demand of senior right holders, the junior or latest issued right holder get the chance to withdraw water. As a result, the senior right holder got more reliable water and rather sharing shortfalls among all, only junior right holders had to suffer from shortage of water (Baker, 2002).

In the Yello River, China long term mean annual flow is measured as 58 billon m3 of which 36 % is reserved for riverine needs and remaining 64 % is allocated among different sectors including agriculture based on entitlement of the users (GIWP, 2007). In Mahaweli River system, Sri Lanka, at start of each season demand form all sectors like water supply, hydropower, agriculture are sent to the Government authority. All these demands are then added with associated losses and environmental needs to estimate the total requirement of water. Depending on rainfall projection for the coming year, several option of allocation to various sectors are prepared and discussed with the stakeholders and finalized version sent for approval to the concern Minister. In the mid of the season, due to fluctuation of climatic condition, rationing or adjustment in allocation is done based on the reservoir storage and rainfall conditions. Hence, allocation may decrease in the mid of the season due to drought condition. The system is largely maintained on previous experience and the users do not have any long- term right over available water resources (ADB, 2009).

2.3 Use of additional hydrological Information:

Assessment of available water resources in a catchment is important part of water allocation decision making process. This task is often supported by predictive and quantitative tools like hydrological model. Hydrological models can be of regional or global. Regional hydrological models work on finer spatial resolution and need to be calibrated with local data (Krysanova, 2017). Moreover, for better performance of regional hydrological model, long term, good quality hydro-meteorological data (i.e. precipitation, temperature, streamflow etc.) is needed which is scarce in many region of the world due to incompleteness of existing data, poor maintenance of instruments and infrastructure and sometimes unwillingness to share exiting data (Masafu, 2016). Whereas global hydrological model function in coarser scale and not needed to be calibrated in local scale. Currently various sources of global data and output of large scale hydrological models are available which can be used as additional hydrological information for quantifying the water resources in absence of conventional hydro- meteorological data (WMO, 2012).

EartH2Observe (Global Earth Observation for Integrated Water Resource Assessment) is an European Union (EU) funded project which integrates large amount of remote-sensing data with ground data and 10 global hydrological/land surface models to produce Global Water Resources Reanalysis (WRR) dataset for more than 30 years period (E2O). The dataset is available in Water Cycle Integrator (WCI) data portal and can be useful for assessing water resources in local and regional scale, determining local hazards like drought or flood risk and can support in decision making process for efficient water management. In the data portal, simulated total daily and monthly runoff is available on 0.5o x 0.5o and 0.25o x 0.25o spatial resolution for a period of more than 30 years. These models are not locally calibrated and forced with a consistent meteorological forcing dataset (WP2, 2017). So, it is very usual that model outputs are associated with unknown bias. In minimizing the effect of bias from a single model, it is recommended to use ensemble mean of output form several global model.

These datasets have been used and evaluated in several studies such as for runoff estimation, flood event occurrences. For instance, Beck et al. (2017) used 966 medium sized catchment observed stream flow data to evaluate daily runoff estimates of six global hydrological models and four land surface models which were produced as part of tier-1 of the EartH2Observe project. They found remarkable difference in performance between models. However, performance of uncalibrated global model was found to be better than uncalibrated local model in snow dominant regions. Linés et al. (2018) used remotely sensed snow coverage data as additional information to investigate its influence in determining availability of water during drought season in Ebro basin, Spain and found better performance which can improve cropping decisions as well as provide higher benefits to the farmer. The usefulness of global data sets can be investigated in water allocation process as well.

CHAPTER 3

Case Study Area

This chapter aims to provide a brief description about the study area that includes catchment, river flow, climate and land use. Moreover, regulation along the river including two major reservoirs are highlighted. The existence of major irrigation districts and their demands are also explained.

3.1 The Murrumbidgee River

3.1.1.Catchment description

The Murrumbidgee River is the 3rd longest river (~1,485 km) in Australia and in the Murray– Darling Basin (MDB), after the Murray (~2,508 km) and Darling (~1,545 km) rivers. The area of the catchment is almost 84,000 km2 which is 8% of the total area of the MDB Basin and provides almost 16% of inflow to the Basin (Burrell, 2017). As shown in Figure 3-1, the region is situated within southern New South Wales and surrounded by Lachlan Catchment to the North, the Murray catchment to the South and West and the Great Dividing Range to the East .

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-1: Location of the Murrumbidgee valley

[ Source : CSIRO, 2008]

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-2: Topography of the Murrumbidgee catchment

[Elevation in eastern part range from over 2,200 metres to less than 50 metres on the western plains (Green, 2011)]

3.1.2.Origin and flow direction

The Murrumbidgee River originates from Fiery Range located in the Australian Alps (Snowy Mountain) which is 1,600 m above sea level and ends at Balranald which is 60m above sea level (Figure 3-2). A figure of longitudinal profile is attached inAppendix C. As shown in Figure 3-3, it flows through steep slopes towards the Tantangara reservoir where part of flow is diverted to the Snowy Mountains Scheme (spread over the Murrumbidgee and the Murry basin) and from there it flows towards the Southeast. Near the city of Cooma, it turns to the North to flow through relatively low elevated area. Afterwards it again changes direction to the Northwest and passes through Australian Capital Territory (ACT) and the national capital, Canberra before joining Goodradigbee and Yass River where the Burrinjuck dam is located. Afterwards, the river flows Westward to Gundagai and meets one of the major tributaries, the Tumut river, which carries significant amount of flow from the Snowy Mountains Scheme and supplies water to storage provided by the Blowering dam at its upstream. Then the river flows Southwest through the cities of Wagga Wagga, Narrandera, Hay and eventually join the Murray River at Balranald. In the middle part of the river, some of the flow leaves the main stream to create the Yanco Creek system. The lower swampy portion of the river is commonly known as Lowbidgee (Burrell, 2017).

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-3: The Murrumbidgee River

(The river from two major dam (Burrinjuck & Blowering) to confluence with the River Murray near Balranald is known as the regulated Murrumbidgee water source) (Green, 2011)

3.1.3.Regulated Murrumbidgee River

The Murrumbidgee River is regulated by fourteen dams and eight large weirs (Figure 3-4). Due to the construction of the series of dam and weir along the river for hydroelectric power generation as well as to fulfil downstream demands, the pattern and volume of flow in the river has changed significantly from the natural pattern (DIPNR, 2004). The portion of the river from Burrinjuck and Blowering Dam to Balranald, where the Murrumbidgee River meets the Murray River is defined as the Murrumbidgee Regulated River (WSP, 2003), which has a length of about 1,200 km. This portion has been declared as the Regulated River by the New South Wales (NSW) Government.

3.1.4.Burrunjuck Dam

Burrinjuck Dam construction was started on the Murrumbidgee River in 1909 and completed in 1928 to store and supply water to fulfil the downstream irrigation demand. It has a surface area of 55 km2 that collects water from an area of 12,953 km2 mainly through the Goodradigbee, Yass and upper Murrumbidgee Rivers. The storage capacity of the dam is 1,026 GL and spillway capacity is 2,506 GL/day. (SW,2009a)

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-4: Major dams and weirs in the regulated Murrumbidgee.

( Source : water.nsw)

3.1.5.Blowering Dam

Blowering Dam construction was started on the Tumut River in 1964 and completed in 1968. Though the catchment area of the dam is only 1,630 km2 that is full of mountains and forest, majority of its inflow is coming through 22 km tunnel of Snowy Mountains Scheme from Lake Eucumbene. Snowy-Tumut Development under Snowy Mountains Hydro-Electric Scheme releases water for generating electricity to meet the winter season electricity demand. Moreover, the stored water is used to meet downstream irrigation demand in the summer. It has a surface area of 44.6 km2. The storage capacity of the dam is 1,628 GL and spillway capacity is 230 GL/day (SW,2009b).

3.1.6.Climate

The climate in Murrumbidgee catchment varies a lot due to the difference in elevation (from 2,200 m to 50 m) (Burrell, 2017) of the area through which it flows. The Eastern highlands have a temperate climate, where summers are warm in most areas and cool in comparatively high altitude areas. There is no dry season in those areas and the winter is very cold. A sub- humid climate prevails on the South-West slopes where the summer is hot. However, a dry semi-arid climate is found in the Western areas with hot summers and cool winters. The hottest months in the area are January and February when temperature vary from 33°C in the West and 16°C at higher altitudes in the East. Winter temperature vary from 3°C to 5°C in the west to 0°C to -2°C in the Eastern highlands (MWRP, 2017).

Rainfall

Rainfall in the catchment is also governed by the elevation. In the higher elevations of the Snowy Mountain area, average annual rainfall is as high as 1,700 mm. On the other hand, in the middle of the catchment it is almost 400 mm and in the western plains only some 300 mm (Figure 3-5). The average annual rainfall in the whole catchment is about 530 mm. The amount of rainfall is almost uniform all around the year (slightly higher in winter season, Figure 3-6) but varies a lot from year to year (Green, 2011). As a result, drought is very common from the mid-section to the west due to irregularities of the occurrence of rainfall as well as rainfall being lower than normal. Thus, it is evident that rainfall in the upper catchment is the main source of water for the river. Moreover, the Eastern highlands, with altitudes ranging from 2200 m to 1500 m are usually covered by snow, which provides an additional supply of water during the warmer weather in the summer by melting (MWRP, 2017). The construction of two major dams at the end of upper catchment provides the opportunity to catch and hold this water and releasing it during the season of peak demand.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-5: Average annual rainfall in The Murrumbidgee Catchment

[Long term average rainfall for the period from 1961 to 1990. ( B urrell, 2017)]

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-6: Average monthly rainfall at Tumbarumba, Wagga Wagga and Hay.

[The loaction of Tumbarumba, Wagga Wagga and Hay is shown in F i gure 3-5. Average rinfall is calculated based on the historical data from the year 1886 to 2016, 1898-2010 &1881 to 2016 for respective location. (Burrell, 2017) (Green, 2011)]

Evaporation

The evaporation pattern, based on Class A pan evaporation observation, is also affected by the differences in elevation. As the elevation decreases from east to west, evaporation reversely increased from 1,000 mm/year in the South-east part to 1,800 mm/year in the Western part (Figure 3-7). It also shows a strong seasonal variation. At Wagga Wagga, in the winter period during July it may be as low as 1 mm/day, while in summer during January it reaches 9 mm/day. (Green, 2011)

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-7: Annual evaporation in the Murrumbidgee catchment

[Class A pan evaporation is calculated based on historical data from the year 1921–1995. (Green, 2011)]

3.1.7.Land Use

Around 520,000 people are dependent on the Murrumbidgee River. This includes the residents of the Australian Capital Territory, the national capital Canberra, which is the largest inland city of NSW; as well as Wagga Wagga and regional cities and towns like Cooma, Tumut, Narrandera, Griffith, Leeton, Hay and Balranald (Burrell, 2017). As shown in Figure 3-8, land use is mostly governed by grazing (64.4%) and agriculture (20.6%). There are commercial forests and conservation land situated in the eastern alpine area of the catchment full of alpine herb fields, moist forests and woodlands. Two large irrigated areas are situated in the middle portion of the catchment known as Murrumbidgee (north) and Coleambally (south) Irrigation Areas (Figure 3-9). Rice, soybeans and corn are grown in summer while wheat, oats and barley are harvested in winter in these irrigated areas (Green, 2011).

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-8: Land use pattern in Murrumbidgee region (Green, 2011)

3.1.8.Irrigated agriculture

Irrigated agriculture is thriving in the region due to the high storage capacity of the two major upstream dams and regulation along the river. It produces 25% of NSW’s vegetables and fruits, 42% of wine grapes and 50% of Australia’s rice. Annual turnover from agriculture is around AUD 1.9 billion, which designates the region as the food bowl of southern Australia (Swainson, 2013). Two major irrigation districts known as Murrumbidgee Irrigation Area (MIA) and Coleambally Irrigation Area (CIA) have been developed in the middle of the region (Figure 3-9). Diversion of water from the river to irrigation canals is done by several weirs to support the irrigators. Additionally, numerous individual irrigation firms also exist along the river that pump water directly from the river (MCIP, 2003).

3.1.9.Murrumbidgee Irrigation Area

The Murrumbidgee Irrigation Area (MIA) occupies 660,000 hectares of land in the northern side (Figure 3-9) of the Murrumbidgee river downstream of Narrandera. It was established by the government in 1912 and is presently operated by Murrumbidgee Irrigation Limited (MI), one of the largest private irrigation companies in Australia. Almost 140,000 ha of land in the area gets irrigation water through the Main Canal and the Sturt Canal that is fed by water diverted from the Murrumbidgee River. The main canal can carry a maximum flow of 6,500 ML/day (75 m3/s) and the Sturt Canal has the capacity to carry a maximum of 1,700 ML/day (20 m3/s). It has over 3,500 km of supply channels and 1,600 km of drainage channels (MI, 2017). Within this area, users hold 647 GL of General Security (34% of total GS) and 297 GL of High Security (75 % of total HS) water entitlements (NSW, 2013).

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-9 Major Irrigation area in the Murrumbidgee region.

(Green, 2011)

3.1.10.Coleambally Irrigation Area

The Coleambally Irrigation Area (CIA) occupies 400,000 hectares of land in the southern side (Figure 3-9) of the Murrumbidgee river upstream of Darlington Point. It was established by government in 1960 and is presently operated by Coleambally Irrigation Co-operative Limited (CICL), the fourth largest irrigation company in Australia owned by its farmer members. Almost 79,000ha of land in the area get intensive irrigation through 518 km of supply channels that are fed by water diverted from the Murrumbidgee River. There are about 491 irrigation farms typically comprising 200 ha of land for each farm within the area that mainly produce rice, wheat, corn, cotton, barley, soy beans, canola along with a variety of fruit and vegetables. Within this area, users hold 391 GL of General Security (21% of total GS) and 10 GL of High Security (3% of total HS) water entitlements (NSW, 2013).

3.1.11.Irrigation demand

Irrigation demand in the MIA and CIA is much higher in the summer than the winter season and varies a lot depending on the total area harvested. As shown in the Figure 3-10, under average climatic condition in the year 2000-01, demand was below 50 GL/month till the end of August. However, demand increases gradually in September but became very high in the month of October which was more than 200 GL. Afterwards, it raised gradually and peaked at 290 GL in January. By the end of cropping season in February, demand again went below 50 GL/month. It was also observed that though rice cultivation covered one-third of the total area, more than half of the demand was consumed by the same crop. It implies that, Rice is the main consumer of irrigation water in the Murrumbidgee region.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-10: Irrigation demand in two major irrigation area (2000-01).

[ Data Source: Khan, 2004 & Khan, 2008]

3.2 Stream flows

Precipitation in the form of rainfall and snow melt from the upstream high lands of the Great Dividing Range (usually snow occurs between June to October) contribute significant amounts of flow to the river. There are some tributary rivers immediately after the dams that also contribute to the river flow (Kingsford, 2001). Flow in the Murrumbidgee river is mostly regulated by two upstream dam. Therefore, river flow is governed by the inflows into those storages and downstream demands. Long term mean annual flows at mid of the river (Wagga Wagga) is 3,732 GL/year. As water demand (mostly from irrigated area) rises during the growing seasons of crops (July to January), flow increases due to high releases from dam whereas reverse is observed in the winter season (Figure 3-11). In addition, due to melting of snow in the highlands of the upper catchment during warm weather in summer, increased flow in upstream of dam also observed. Moreover, As the water is diverted to MIA through main canal in between Narrandera and Wagga Wagga and to CIA, Sturt Canal and several distributaries downstream of Narrandera, discharges declined consequently.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-11 : Long term mean monthly river flow at different location

(GL/month)

( Yass and Goodradigbee is located at the upstream of Burrinjuck dam and other staions are located at the downstream of the dam. During the irrigation season, river flow become higher due to the high release from the dam to meet downstream irrigation demand)(data source : http://realtimedata.water.nsw.gov.au)

The existence of the Snowy Mountain schemes upstream of the Blowering dam has increased the reliability of water supply for the downstream irrigation demand, even in the period of low rainfall or drought. Usually the scheme contribute almost 25% of total flow to the river, but this may increase up to 60% of the total flow in the dry season (Snowy Water Inquiry, 1998).

Daily and annual river flow from 1899 to 2018 at Wagga Wagga is shown in the Figure 3-12 and the long-term mean daily flow in different locations is shown in Figure 3-13. It is evident from the figures that despite high regulation along the river, large variations in river flow happen from year to year.

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-12 : Daily and annual discharge at Wagga Wagga from 1899 to 2018.

(data source : http://realtimedata.water.nsw.gov.au)

Abbildung in dieser Leseprobe nicht enthalten

Figure 3-13 Mean daily flow at different location along the Murrumbidgee river.

(Green, 2011)

CHAPTER 4

Water Allocation in the Murrumbidgee

This chapter provides a brief description about the water sharing process in the Murrumbidgee. It also describes various license category, their priority, entitlement and the rules followed in available water determination (AWD) to allocate water to different categories. Water sharing for environmental purposes and carry over rules are also explained. It comprises of an analysis of sequential process followed in water allocation with its components.

4.1 Water Sharing

Water in the Murrumbidgee Regulated River is shared between the users in accordance with the statutory Water Sharing Plan for the Murrumbidgee Regulated River Water Source, 2016 (previous approved version was published on 2003 but has since been suspended by the 2016 version). The plan includes rules for extraction of water from the river to ensure a balance between environmental needs and regional demand of the community (DWE, 2009).

4.1.1.Basic water rights and license

It is necessary to get proper approval from the government authority to extract water from the Murrumbidgee regulated river. This approval generally depends on the purpose of use, location of the users. If a user owns land on the bank of the river that has direct access to the river (Figure 4-1), he can extract certain amount of water from the river for non-commercial uses under basic landholder rights to fulfil basic domestic requirements like washing, watering domestic animals, gardening etc. However, for other purposes, users need to get a Water Access Licenses (WAL) from the authority. A sample of WAL in NSW region is attached inAppendix B.A WAL is issued for purposes like local water utility, town water supply, domestic needs, stock watering, irrigation or industrial use. Each water access license (WAL) is composed of a share and an extraction component. The share component describes the amount of water that can be taken from a source like the Murrumbidgee Regulated River whereas extraction component specifies times, rates and circumstances of extraction of water. For irrigation, WAL can be of two types, which are known as a (i) High Security WAL, and (ii) General Security WAL and are permanent in nature. This means these do not need to be renewed every year. An irrigator can extract a certain amount of water from the river based on the volume that has been mention in his license, which is known as the entitlement volume or share component. Usually, irrigators having larger farm area got more entitlements of water to use. Once an irrigator gets a license, he is entitled to extract water unless the authority cease the license for some reason. Issuance of new WALs for irrigation has been stopped in 1986 to improve the balance between supply and demand (Turral, 2005).

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-1: Condition to extract water by Domestic and stock right.

(Source: water.nsw.gov.au)

4.1.2.Entitlement of License categories and allocation

There are various categories of water access licenses (WALs), each with different purpose of usage and priority for allocation and extraction. For instance, domestic and stock WAL are for those users who don’t have direct access to the river, but need water for domestic use and watering of stock and pets. Local water utilities WAL are for the purpose of providing water supply and sewerage services to non-metropolitan urban communities, town water supply WAL are for the purpose of water supply to communities for domestic consumption and commercial activities in metropolitan areas. High Security (HS) WAL licenses are traditionally held by irrigators with permanent plantings (e.g. horticulture, grapes, citrus), high value crops where the investment is very high and that require a constant annual application of water to remain productive. General Security (GS) licenses are mostly held by irrigators growing annual crops such as rice, wheat, corn, cotton, vegetables and livestock. The price of a unit volume of water for HS is much higher than for GS due to the higher reliability to receive the water allocation. For instance, in 2017-18, the price of each ML of GS water was 3.06 AUD whereas HS was 5.96 AUD.

The volume of water that a user can extract from the river is mentioned in his WAL and known as an entitlement. The total entitlement for specific purposes like local water utility, domestic and stock usage are expressed as total volume of water for each year (e.g 23,403 ML per year). On the other hand, entitlement for general security, high security, conveyances are calculated based on the number of unit shares that are held in the license. In general, each unit share authorizes for maximum 1 mega liter/year of water extraction from the river. So the higher the quantity of unit shares held in a license, the higher is the volume of the entitlement for use. For instance, if a general security access license holder has 300 unit shares in his license account, his entitlement to extract water is a maximum 300 mega liter/year (=300 unit share * 1 mega liter/share/year). There are 1,891,815 nos. of unit shares issued under general security license category, so total entitlement under this category is 1,891,815 ML/year.

In reality, an entitlement cannot give assurance of the volume of water to be available for use in a certain year. Rather, it depends on the allocation announcement, which is based on the availability of water. If the total available water in a year is more or equal to the total entitlement of all license categories, every license holder will get an allocation equal to entitlement volume. In contrast, if the available water is less than the total entitlement, water is allocated within the license categories based on their priority mentioned in the WAL. As a result, it is very usual that low priority categories will get only a percentage of the entitlement as allocation. For instance, if total number of license of all categories are 2,500 which provide entitlements of 2,500 ML (1 ML for each license) and in a certain year total available water is calculated as 2,000 ML then the shortfall will be 500 ML (=2,500-2,000). As a result, some low priority license holder will get a lesser amount of water as allocation than their entitlement volume. Hence, the entitlement can be defined as the maximum amount of water a license can authorize its holder to extract water from the river if sufficient water is available in the system. On the other hand, allocation authorizes a user to extract water from river in a certain year which can be less than or equal to entitlement. An analogy is that entitlement is the size of the water user’s ‘bucket’ and allocation is the amount of water to be placed in that ‘bucket’ in a given year (Figure 4-2). So, allocation may change year to year whereas entitlement is fixed against a license.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-2: Entitlement and allocation for Water Access Licenses (WALs).

High priority license holder get water allocation before allocation is assigned to general security licenses. For instance, if total entitlement of high priority categories are 800 ML and low priority categories are 2,000 ML and in a certain year amount of available water is found to be 2,200 ML then after fulfilling requirement of all high priority categories (e.g 800 ML), remaining 1,400 (2,200-800) volume of water will be distributed to lower priority categories. Various categories of WAL, their priority and total amount of water entitlement for each categories are shown in Table 4-1.

Table 4-1 Total Entitlement for Access Licenses in the Murrumbidgee Regulated River

Abbildung in dieser Leseprobe nicht enthalten

* only when flows are surplus in stream. Data source: (WSP, 2016) & (Green, 2011)

4.1.3.Available water determination (AWD)

In the Murrumbidgee region, at the beginning of each water year (1st July) which spans from 1st July to 30th June of the following year, the NSW office of water determines the minimum assured volume of water that will available for allocation to the licensed user in the coming year. First, the water stored in all storage at the start of the year are measured. Then the expected minimum natural inflow into the storage and releases from Snowy Hydro Limited into the Blowering dam in the coming year are calculated from historical data. By adding this expected inflow volume with storage water, the total water that will be available in coming year into the system are determined. As the future inflow estimation is based on the historical minimum natural inflow into the storages, this calculation is considered as very conservative. From this volume, various losses (i.e evaporation, transmission), environmental need of water, earlier commitment (i.e carryover), end of system flows and minimum reserves for storage are deducted to calculate the total volume of water that can be delivered to the licensed user. This volume of water is known as available water for consumptive use or accessible water for the next 12 months (Figure 4-3).

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-3: Computation of available water for consumptive use in the Murrumbidgee regulated river.

From this volume of water, allocation to each license category is given for whole year based on their priority. This is done by an official announcement which is knows as Available Water Determinations (AWD) for each license category and initial AWD announcement is given on 1st July of each year. After the initial AWD announcement, each license holder gets the allocated water in his account. For general security and high security license categories, the AWD express the amount of water that will be available for each unit share. As described earlier, each unit share is composed of 1 ML of water. So the AWD express an amount between 0 to 1 ML for each share for whole year. For instance, if a user has 1,000 unit share in his GS access license, it will provide entitlement of 1,000 mega liter (ML) water per year. Again if AWD specifies 0.3 ML/unit share allocation in a certain year, the user will actually receive 300 ML (0.3 ML x 1,000 unit share) in their account for that year (Irrigation Fact, MI). Moreover, it also means that the user will get 30% of his total entitlement. AWD is made for each license categories, so 0.3 ML/unit share allocation for GS license category means all GS license holders will get the same amount of water for one unit share. In the Murrumbidgee, 1,891,815 unit shares exists under the GS license category which provide 1,891,815 ML of entitlement/year, so in the example, the total allocation to the GS category will be 567,545 ML (30 % of 1,891,815 ML entailment).

4.1.4.Revision of the AWD

Typically, the quantity of available water for allocation increases as the year progresses. This is due to the fact that most of the time inflow into the storages exceeds the expected inflow volume that was taken into account initially. Usually inflow/resources into the storages increases due to increased rainfall or lower transmission losses than expected. As a result, new AWD is made to allocate this extra water to the license categories which had a shortfall in allocation. Moreover, after new AWD, license categories with new allocation get this new water added with old allocation into the license holder’s accounts. If allocation to a license category reaches its full entitlement, no new allocation is made for that category. Hence, revision in AWD continues until every license category gets an allocation of 100 percent of entitlement. The NSW office of water regularly asses the inflow condition and generally makes revision in the AWD every 15 days during wetter condition or very critical times of the year. Usually, as time passes, the frequency of AWD revision decreases until the end of season (February). After the end of February, any improvement in inflow condition is usually kept reserved for the upcoming year’s allocation.

4.1.5.Supply of allocated water

Every water access license has an extraction component where limits on the times, rate and circumstances of extraction are usually mentioned (DPI Water, 2015a). Moreover, each license has a water allocation account where water is credited according to the allocation volume declared by AWD announcements (Ribbons, 2009). A user have to place a water order to the authority mentioning the amount of water need to extract, location and rate of extraction, intended use and other information. In headwater dam, the operators release the volume of water summing all water orders and it takes 5-7 days to reach the irrigation districts. When the user extract water according to the order, same volume of water is deducted from his allocation account. In this way, record of water availability and usage is maintained for every license. At the end of the year (30th June), if there is any water remain unused in the account, it will be forfeit except general security and conveyance category.

4.2 Water allocation rules

In the Murrumbidgee region, water is allocated to different license categories according to their priorities as specified in Water Sharing Plan for the Murrumbidgee Regulated River Water Source. It ensures that when water is sufficient, everyone will get the required amount of water. In contrast, when water is low, everyone gets water according to their priority. Before making AWD for the license categories, a sufficient amount of water is set aside for minimum storage reserves, environmental needs and losses. Water for critical human needs (domestic usage, town and local utility supply) gets higher priority than commercial purposes during water allocation. Requirements for these high priority needs are comparatively lower than the low priority needs such as general security requirements. The priority of WAL categories for water allocation and entitlement percentage with respect to total entitlement in the catchment is shown in Figure 4-4. As illustrated, domestic, stock, town and local water has the highest priority for getting water with only 3% of total entitlement whereas general security has lowest priority but has highest entitlement of 69%. As a result, allocating full entitlement to the critical human needs is usually done at the initial AWD even in severe drought condition.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-4: Priority and entitlement of license categories

Data source: (WSP, 2016) & (Green, 2011)

4.2.1.Environmental Water

In the water sharing plan, detailed rules for preserving environmental water from allocation for WAL categories are described. In the plan, long-term average annual flow in the Murrumbidgee river system is calculated as 4,360 GL/year and average extraction for consumptive use is set at a maximum of 1, 925 GL (WSP, 2016). This means almost 56% of total annual flow volume is given for the maintenance of basic ecosystem health dependent on the river. Annual flow into the system includes inflows into the two upstream major reservoir as well as the tributaries inflow into the downstream of the dams. There are mainly five types of provision for environmental water, which are known as end of system flows (EOS), dam transparency, dam translucency, environmental water allowances and provisional storage. End of System Flows (EOS)

Provision of End of System Flows is to ensure connectivity downstream of major extraction points of the Murrumbidgee River to the junction with the Murray River. It will ensure the continuity of ecosystems along the river reach. According to this rule, the minimum daily flow in the river at the Balranald gauge (ML/day) = 300 + [(0.4 * 95th percentile natural daily flow for the month)-300]. The 95th percentile natural daily flow for the month is the flow that exceeds 95% days in a certain month without regulation generated by computer model. In addition to this, a minimum flow of 50 ML/Day is required at Darlot Gauge in the Billabong creek.

Dam Transparency

Dam transparency rule is provided to ensure that downstream of the river never dries up even in very low inflow conditions. This ensure continuous release of water from both Burrinjuck and Blowering dam. It includes a rule that, if the inflow into Burrinjuck is less than 615 ML/day and Blowering is less than 560 ML/day, all the inflow in the previous 24 hours should be released from both dams in that day. However, if the inflow is greater or equal to this threshold, the daily release will also be equal to that threshold value, thus storing the difference for later use.

Dam Translucency

Construction of dam in the river restricts the natural flow variability in the downstream part of the river. Hence, to mimic the flow variability in the downstream of Burrinjuck dam, when inflow in excess of 615 ML/day are observed between 22 April to 21 October, a percentage of inflow is released form the dam. This is known as Dam Translucency rule, which is only applicable for specific days in Burrinjuck dam.

Environmental water allowances

Environmental water allowance (EWA) is the volume of water that is kept in the dam and released afterwards to create minor flooding events in the downstream for the benefit of wetland inundation, fish passage, bird and fish breeding, water quality and water dependent Aboriginal cultural values. There are three types of account for this water, which are known as EWA1, EWA2 & EWA3. When allocation to the general security exceeds 60% of total entitlement, 50 GL of water is set aside and credited to EWA1 account. EWA2 & EWA3 are credited with water according to the volume of transparent and translucent releases, and when allocation to GS exceeds 80% of entitlement.

Provisional storage volume (PSV)

Provisional storage volume (PSV) provides the scope to retain a certain volume of water in the storage to increase probability of spill events at the beginning of the following year which will enhance the probability of natural flows in the downstream part of the river. There are two types of account for this water which are known as PSV1 & PSV2. When allocation to the general security exceeds 60% of total entitlement, 25 GL of water is set aside and credited to PSV1 account. Moreover, if GS allocation equals 80% of total entitlement, for further each 1% increase in GS allocation, 8.75 GL have to be credited to PSV1. PSV2 get water in accordance with the increase in EWA3 account.

4.2.2.Allocation to Water Access License (WAL) categories

According to water sharing plan, water is allocated to different categories of access license account through Available Water Determination (AWD). AWD specifies the amount of water each category will get in a given year and generally varies from year to year due to the availability of water assets. The rules for water allocation is described below as per hierarchy in priority:

Specific Purpose Licenses

Domestic and stock, local water utility license category get 100% of entitlement volume as allocation at the commencement of every year before any category get water. Afterwards, aboriginal cultural, research and town water supply categories get 100% of entitlement if required volume of water is found available. Allocation to aboriginal cultural WAL is given for water needed in cultural and spiritual events of aboriginal peoples living in the region. Local aboriginal people need to apply for the required amount of water to the NSW government authority mentioning the purposes of the event with prior approval from Local Aboriginal Land Council. If a researcher needs water for experimental purposes, they can apply for water allocated against research WAL.

High Security access license

At the commencement, 95% of total entitlement volume should be given to the high security WAL category. In a given year, if the general security (GS) allocation is more than 94% then for every 1% further increase in GS allocation, an increase of 1% should also be made to HS in addition to the already allocated 95% until 100% of the HS entitlement is reached.

Conveyance

The conveyance water is divided into three categories which are known as Murrumbidgee irrigation, Coleambally irrigation and regulated river. Conveyance water license are issued to the irrigation authority considering the intended loss (channel leakage, evaporation) occurred to transport water to the user’s location through irrigation canal. As an example, when a farmer submit an order for 200 ML of water to the irrigation authority, the dam operator will release more than 200 ML, keeping the intended losses from dam to the user’s location in mind. The price of this extra water is included in the service charge of the irrigation authority that is paid by the users. Allocation to these categories increases with the increment of allocation to GS and HS category as operational loss increases with the increase in delivery of water.

For Murrumbidgee irrigation (conveyance), 100% of entitlement is allocated when GS allocation exceeds 60% of entitlement. However, if GS is less than 20% (0.2ML/share), 550 ML for each 1% of GS should be allocated to this category. A further 550ML for each 1% of HS is given to this category until the allocation to HS reaches to 95% of its entitlement. An additional 1650 ML for each 1% of GS should be allocated when GS allocation exceeds 20% but not more than 50% of entitlement. An extra 3200 ML for each 1% of GS is added to this category when GS allocation crosses 50% until reaching 60% of total entitlement.

For Coleambally irrigation (conveyance), 100% of entitlement is allocated when GS allocation reaches to 100%. However, if GS allocation is less than 35% (0.3 ML/share) of entitlement, the allocation will be 111,600 ML. When GS allocation crosses 35% of entitlement, 760 ML for each 1% is added until GS cross 40%. From 40% to less than 100% GS allocation, this category gets additional 243.3 ML for each 1% GS allocation.

General Security access license

AWD for general security license category are made when water remains available (Figure 4-5) after establishing the allocation to environmental water, full entitlement of domestic and stock rights and licenses, full entitlement of local utilities, aboriginal culture, research, town water supply, 95% of high security entitlement, sufficient water for loss in delivery of irrigation water by evaporation and channel leakage (conveyance) and unused water from previous year known as carryover.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-5: Water allocation simplified.

( General security license holders get water after high priority users and system operation requirements are met.) Data source: (WSP, 2016)

Supplementary Water

When water becomes excess due to heavy rainfall or from the spill of dams, license holder of this category can take water from the river. The NSW government announces this excess water situation, and only at that time can the license holder extract water under this license. As the water in this license is only available during unpredictable wet conditions, full allocation is made at the commencement of each year and it is excluded from the total available water for consumptive use.

4.2.3.Carryover

In the Murrumbidgee regulated river, if a license holder does not extract all of his allocation from the river before the end of the water year (i.e 30th June) then the unused water will be forfeited, or lost. However, general security and conveyance license holder can carry over the unused allocation to the next year to a volume of not more than 30% of entitlement. This provides flexibility to the license holders to use and hold certain amount of their allocation into the storages and make it available in the subsequent year confirming minimum required volume of water at the start of each year. It also reduces the ‘use it or lose it’ mentality and allows users to manage their water efficiently, especially during dry years. However, carryover affects the amount of new water that will come into an account at initial AWD as no license holder can have more than 100% of their entitlement. As a result, AWD for these categories are the sum of new water and the carryover volume from previous year’s unused water. For instance, as shown in Figure 4-6, suppose a general security license holder has entitlement of 1000 ML/year and got allocation of 700 ML (0.7 ML/share or 70% of entitlement) in a certain water year. But he has used only 200 ML of water in that year and 500 ML remain unused at the end of the year. He can transfer maximum 30% of total entitlement which will be 300 ML (30% of 1000 ML/year) to the next water year and remaining 200 ML (unused, 500 ML – carryover, 300 GL) will be forfeited or lost. In the coming year, if initial AWD allocate the same license holder 0.2 ML/share or 20% of his entitlement (= 200 ML) then practically he will have total 500 ML ( initial AWD, 200 ML + carryover from previous year, 300 GL) of water initially for the rest of the year.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-6: Carryover rules for General Security and conveyance license.

4.3 Water allocation process

In determining the total volume of water that will be available for allocation at the beginning of each water year, calculation process shown in Figure 4-7 is usually followed. Each part of calculation is comprises of several components among which 2 to 3 components contribute major portion of water. After analyzing the available historical data from 2011 to 2016, water stored in two major reservoirs prior to 1st July of each year was found to contribute 98-99% of all stored water that is taken into consideration (Part-A). In determining the future increase of water assets, expected inflow into two major storages provide nearly 90% of total expected water (Part-B). To omit the volume of water that could not be accessible from the available water for allocation in whole year, transmission losses (i.e. seepage, evaporation), water required to maintain connectivity from start to end of regulated river are taken as major factor which usually account for 75% of total volume (Part-C). Therefore, Total accessible water is the sum of stored water and expected inflow less the expected losses/reserves for whole year.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-7: Annual accessible water calculation process.

(Total accessible water is the sum of stored water and expected inflow less the expected losses/reserves for whole year) Data source: (Burrell, 2017)

Calculated total accessible water is then distributed in two steps. In the 1st step, water is allocated to existing liabilities and the remaining water is allocated to the access license accounts in the 2nd step as shown in Figure 4-8. Water liabilities consist of water that is due from the previous year and has to be delivered in the current year. Surface water carryover is the majority of liabilities and consists of allocation to general security and conveyance categories. The volume of water required for environmental purposes are calculated based on water sharing rules (2003, 2016). Shortfall of released environmental water in a given year has to be released in the next year, which is also considered as a part of the liabilities. After considering the water set aside as liability, the remaining water is distributed to the licensed users based on priority as described in section 4.2.

Abbildung in dieser Leseprobe nicht enthalten

Figure 4-8: Available water determination (AWD).

(Total accessible water is allocated to each category based on their priority) Data source: (Burrell, 2017)

CHAPTER 5

Methodology

This chapter provides a description of research methodology and techniques used to achieve the research objectives and address the research questions. It comprises data collection activities; development and principals of allocation model; development of framework for measuring consistency of water allocation decision process and additional hydrological data to assess the goodness of their use in the allocation process.

5.1 Research Design

In preparing the research proposal, it was realized that the proposed research is unique in nature. Hence, to address the problem stated in section 1.2, thorough understanding of water allocation process in the Murrumbidgee region was felt to be necessary which requires collection and analysis of relevant data. Furthermore, as there was no funding for field work, the work was fully dependent on the online data sources. So, the research was designed considering the uncertainties associated with the availability of required and useful data. Under this circumstances, the research was divided into three phases (Figure 5-1) stated below:

Phase 1: Study and data collection

In the initial phase, all efforts were given to understand the surface water sharing process in the Murrumbidgee region. In doing this relevant literature and reports were studied. Moreover, hydrological data, water allocation data, irrigation and cropping data was collected and analyzed. This phase was crucial as the following phases are dependent on the outcome of this phase. Where required, additional data collection and analysis was also done in other phases to supplement the research outcome.

Phase 2: Model and framework development

In the second phase, an allocation model using Microsoft Excel was developed to emulate the past water allocation decision process using the data collected in previous phase. Various trials were done to explore and evaluate dependency of parameters like reservoir level, inflow into the reservoirs in the decision making process were evaluated. A framework to assess the consistency of water allocation to General Security entitlement was also developed in this phase.

Phase 3: Evaluation of decision making using additional hydrological information

In the final phase, reanalysis data obtained from the EartH2Observe (www.earth2observe.eu) project was used in the model that was setup in phase 2 and simulation of water allocation to GS was done. The result was evaluated using the same framework developed in phase 2 to assess the consistency of decision making process.

5.1.1.Activity Diagram

Different activities to perform this research is presented in the following activity diagram:

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-1: Sequence of activities followed in the research

( Color codes differentiates the pahse of activities)

5.2 Water sharing rules and allocation process

A detail study of the Murrumbidgee catchment has been done in the initial stage of this research. Online contents and reports can be found in the website of NSW office of water, WaterNSW, Commonwealth Scientific and Industrial Research Organization (CSIRO), Murrumbidgee irrigation and Murray-Darling Basin Authority has been studied. Scientific papers on various issues in the Murrumbidgee catchment were also studied to assess relevancy with this research. Brief description of water sharing rules and allocation process is described in CHAPTER 4.

5.3 Data Collection

Water allocation data for 2004-05 to 2014-15 has been collected from the online resources of NSW office of water (www.water.nsw.gov.au) and 2015-16 to 2016-17 was collected form the General Purpose Water Accounting report 2015–2016 published in the same source. Due to the availability of relevant data for these limited periods, the study is also confined within these years. A sample of annual allocation announcement data is attached in tabular from in theAppendix A. Critical water planning for the Murrumbidgee Valley (Issue 1 to 38), available water determination, & General Purpose Water Accounting report from 2011 to 2017 was also collected from same source and studied to get information regarding the initial and consecutive revision in allocation process. Daily water storage, inflow volume in Burrinjuck and Blowering reservoir were collected from the realtimedata.water.nsw provided by NSW office of water. Daily outflow from the reservoir were collected from St. no. 410008 and 410073 for Burrinjuck and Blowering respectively. Daily reservoir inflow are collected from the St. no. 410131 and 410102. Details of the stations and period of data availability is shown in Table 5-1. Using daily reservoir storage and outflow data, daily inflow into the reservoir was calculated with mass balance approach using equation (5-1), in case of missing or unrealistic inflow data. Daily storage and release volume (01/07/2007 to 01/07/2017) from the Blowering is shown in Figure 5-2 and Burrinjuck is shown in Figure 5-3. As can be seen, release volume from both storages increases in the cropping season (October-February) and during winter the reservoir storage start to raise due to high inflow and low release.

Table 5-1: Details of data acquisition station

Abbildung in dieser Leseprobe nicht enthalten

Net Inflow (I) = Outflow (O) + change in storages volume (ΔS) (5-1)

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-2: Blowering daily storage and outflow (July 2007 to July 2017). [ B lue line continuously trace the volume of water stored in the Blowering reservoirs and red column chart shows the daily release of water from the reservoir (Data source : water.nsw)]

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-3: Burrinjuck daily storage and outflow (July 2007 to July 2017). [ B lue line continuously trace the volume of water stored in the Blowering reservoirs and red column chart shows the daily release of water from the reservoir (Data source : water.nsw).]

5.4 Water allocation model

A water allocation model for the Murrumbidgee region to simulate the reservoir operation and allocation to various WALs has been developed using Microsoft Excel. The objective of the model is to emulate the GS allocation process that have been observed in the previous years. In doing this, as required data were available from 2011-12 to 2016-17, the model can simulate the water allocation for this period.

5.4.1.Input data

In developing the model, various component in making AWD has been collected from GPWAR reports for the year 2011-12 to 2016-17. All the data are provided inAppendix D. This data is essential to emulate the water allocation process. Daily observed inflow into the reservoirs are used for this simulation period. Rules for allocating the water to various categories and environmental requirement are also provided to simulate the water allocation based on priority.

5.4.2.Model setup:

The developed model can be divided into two parts; one is resource calculation part and another is AWD part (Figure 5-4). In the resource calculation part, water that will be available for allocation for the whole year in each time step (one day) is determined. Upon the request of the operator, available water is sent to the AWD part where the model allocates water to the different categories and sends back the volume allocated to each category in the resource calculation part where water is released from the available resources (Figure 5-5) according to the total allocated water and demand in each time step. After releasing water according to the demand of different category, every time step have new resource balance for allocation. Resource calculation part of the model run on daily basis and each day is defined as one time step. On the other hand, AWD for different categories is done on the date when it is required by the operator.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-4: Model workflow

5.4.3.Resource calculation

As an input into the model, daily observed inflow into Blowering and Burrinjuck reservoir, water stored in all storage, expected inflows, predicted losses and water liability at the start of each year for 2011 to 2016 is given. The basic formulas that are applied in the model are as follows:

Cumulative volume of water for allocation at any time step “n” for 365 days or 1 year will be,

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-5: Reservior operation in each time step

To calculate the expected inflow into the reservoir, the long term monthly average inflow (Figure 5-6) into the Murrumbidgee system (NSW, 2010) has been used to mimic the flow variation in different month. Hence, the total inflow expected in a day of a certain month was calculated based on the following equation:

Expected inflow into the reservoirs in any time step n, Iexn = Itot x Inp; (5-4)

Abbildung in dieser Leseprobe nicht enthalten

So, total expected inflow for remaining days (365-n)of the year at any time step n will be,

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-6: Average inflow and demand curve used in the model setup.

( Long term average monthly natural inflows (monthly total and %) into the Murrumbidgee system (NSW, 2010) and monthly water requirements (GL) (monthly total and %) in the MIA & CIA for 2000/2001 crops under average climate conditions (Khan, 2004 & 2008)

Moreover, daily releases of allocated water from the reservoir to meet irrigation demand (GS, HS & conveyance) were calculated in the model using irrigation demand observed in the MIA (Khan, 2008) and CIA (Khan, 2004) in the year 2000-01 as shown in Figure 5-6. As the demand is unknown for each year, so total allocation volume is taken as the total demand. Moreover, as the monthly demand for both irrigation area in average climatic condition is available for this year, so it is assumed that it can be useful to use in other years. It is to be noted that the demand in each day is calculated as the percentage of total demand (allocation). So, the percentage value will be same for a given month in every year, but the release volume will change depending on the allocation volume or demand. Release volume for losses and other demand like domestic, town water, environment etc are taken as equal for every day all around the year. When the calculated volume in any time step in the reservoirs exceeds the total storage capacity, the extra volume of water will spill from the reservoirs and the spill volume will be the difference between calculated storage volume and storage capacity (2,657.74 GL) of the reservoirs. As a result, the storage volume in that time step will be corrected after deducting the spill volume. Total volume of water released or calculated outflow from the reservoirs in a time step n will be,

Routn = Lon + Raun + Rspn;

And expected loss requirement for remaining days (i.e 365-n) at time step n will be,

Abbildung in dieser Leseprobe nicht enthalten

In determining the release volume from the reservoirs, total allocation that exist in that time step are taken into consideration. A schematic diagram to illustrate the calculation of total volume of water available for allocation in any time step is shown in Figure 5-7.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-7: Cumulative volume of water for allocation in the model.

(In each time step, this calculation is done by the model and final volume is sent to the AWD model if required by the operator to make calculation for allocation with new values)

5.4.4.Available water determination for license categories (AWD)

This part of the model starts with the cumulative volume of water for allocation for whole year at any time step “n” that is determined in the resource calculation part. As input, total entitlement of each license categories and rules for allocation are given so that allocation volumes cannot exceed entitlement volumes. A flow chart of this part of the model is shown in Figure 5-8.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-8: Available water determination for license categories in the model

At the start, water assets available for allocation and water liabilities is given as input into the model. If the available water asset (A) is greater than zero then it move forward to read the value of water liabilities (B). If water asset (A) can fulfill the total requirement of water liabilities (B), it move forward, otherwise the model stop to calculate further as there is no water available to fulfil further demand. By moving forward it allocates water based on the entitlement of very high priority license categories like Town and local water utility, Domestic and Stock, Basic Rights (D). If more water is available then the model allocates water to High security (HS) category entitlements. As per water sharing rule, even though sufficient water is available to fulfill the total entitlement of HS, after allocating a maximum of 95% of the total HS entitlement (F), the remaining water is held for allocation to General security and conveyance. However, if the remaining water is more than 60% of the total entitlement of (GS+ Conveyance), 75GL of water is kept aside as environmental requirement (EWA1) and provisional storage volume (PSV1) reserve (H). After deducting this water, if the remaining water is found to be more than 95 % of the total entitlement of (GS+ conveyance), HS should be given 5% of remaining entitlement (J) which will fulfill 100% of total HS entitlement. Afterwards, remaining water will be given to the GS and Conveyance (L).

5.4.5.Model performance evaluation

To evaluate the model performance in simulating GS allocation process, three matrices are used: Pearson’s correlation coefficient (r ), Percent bias (Pbias) and Kling–Gupta efficiency (KGE).

Pearson’s correlation coefficient (r )

It describes the degree of linear relationship between simulated and observed data. The value ranges from -1 to 1. If the value goes close to zero, correlation become very low. On the other hand, if it shows value close to -1 or +1, relationship become very good (positive or negative) (Moriasi, 2007). Pearson’s correlation coefficient (r ) is calculated with equation (5-8).

Abbildung in dieser Leseprobe nicht enthalten

Percent bias (Pbias)

It indicates the average tendency of the simulated values to be larger (overestimate) or smaller (underestimate) than the observed data and expressed as percentage. The perfect value of Pbias is zero. Hence, value close to zero indicate accuracy of the model whereas positive indicate overestimation and negative indicates underestimation (López, 2017) than observed data. Pbias is calculated with equation (5-9).

Abbildung in dieser Leseprobe nicht enthalten

Kling–Gupta efficiency (KGE)

It measures bias, variability and correlation simultaneously. It can avoid the problems of high sensitivity to extreme values and bias. KGE value ranges from – α to 1. Value close to 1 indicates perfect simulation (López, 2017). KGE is calculated with equation (5-10).

Abbildung in dieser Leseprobe nicht enthalten

5.5 Consistency in water allocation decision making

As we discussed in earlier chapters, GS allocation decision revised several time in a year unless the allocation reach to total entitlement (100%) before the end of irrigation season. It is also observed that number of revisions and difference between the initial and final allocation varies between years. As cropping decision largely depends on the early allocation decision before planting season, it is desirable that allocation to GS remain same from the start to the end of the year. If the allocation decision changes from the initial one, more the change happen, less will be the consistency. In contrast, if the frequency, timing and extent of revision in water allocation is less, more will be the consistency. For instance, suppose in a year decision makers change the allocation for 10 times and in other year for 2 times and difference between final and initial allocation is 50% and later one is 20%. In general, we define 1st year less consistent than the later year as former year observed significant revisions both in number and extent.

5.5.1.Framework for consistency assessment

It is realized that a framework is necessary to measure the consistency of GS water allocation decision for different years in the same platform. But there is no measurement tools exists to quantify this. Therefore, a framework is prepared in this research considering graphical representation.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-9: Measurement of consistency for two allocation scenario

As shown in Figure 5-9, two allocation scenarios are shown with two allocation curve, Curve1 (green) and Curve2 (dotted red), where both allocation reaches the same quantity at the same time with same initial allocation. However, Curve1 experience 3 revision and provide more quantity of allocation in early time whereas Curve2 experience 4 revision and give less allocation in the early time. So, we define Curve1 shows better consistency than Curve2. To Quantify this, as we can see the area under Curve1 (green dashed) is more than area under Curve2 (pink filled area) [(A+B1)>(A+B2)], we propose to take this area as an indicator of consistency. Hence, more the area under allocation curve more the consistency. Moreover, we can extend an imaginary line from the highest allocation point as if the allocation was constant from the beginning of the year (ideal allocation scenario) and compute the area under that line which will be same for both cases[(A+B1+C1)=(A+B2+C2)]. To compare it with ideal scenario, we can divide it by the total area under perfect allocation curve and name it Consistency Index (CI). It will also make the index value unit less.

5.5.2.Hypothesis regarding best and worst scenario

To measure the consistency, it is needed to describe the best scenario and the worst. It is obvious that if the allocation decision don’t change from the initial allocation and remain same until end of season, the consistency of the decision process can be define as best (Figure 5-10). This scenario will become true if the predicted volume of available water for allocation in a given year never decreased than the actual observed volume all through the year. In that case, at the very beginning of a year, farmers can plant maximum amount of crops based on the crop- water requirements and benefit will also be higher. However, it is difficult to define the worst scenario. Because even if zero allocation prevails all through the year, still it can be considered as consistent decision as allocation never change. In reality, the worst will happen if zero allocation prevails nearly all through the year and full entitlement allocation given at the end of the irrigation season. In that case, irrigators will not have any water to meet cropping demand but at the end of the cropping season, full entitlement of allocation will be useless as cropping season will end. However, they can carry over it to the next year but limited to 30% of entitlement. Based on this hypothesis, a framework is developed and sequence of calculation is described in the following section.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-10: Hypothesis on consistency measurement

5.5.3.Step-1: Representing allocation decision in a step diagram

First of all, all water allocation decision (ML/share) are converted into the percent of entitlement by dividing the cumulative allocation volume by total entitlement volume. Then the allocation percentage are plotted against the allocation announcement dates. As the allocation volume remain constant between two consecutive allocation announcement dates, the graph will be straight in between them but change will happen when allocation volume changes. As a result, the curve will be a step line graph which can be defined as Allocation Curve for that year.

Example:

In the year 2013-14, the initial GS allocation (1st July) was 681 GL for the whole year and total entitlement was 1892 GL. On 01-Aug-2013, 132 GL of water was added with the previous allocation of 681 GL water. Therefore, due to the change in annual allocation decision, on that day the allocation was reached to 813 GL (= 681+132) for whole year. On 15th august, the allocation was revised again and 57 GL of water is added with the previously announced 813 GL and resulting total allocation reached to 870 GL (=813+57). The decisions has been changed further for several time and on 15-Apr-2014 the final allocation was announced as 1532 GL which did not change further and remain fixed until end of the year (30-Jun-2014). In this way, total allocation on each announcement date were converted into the percent of total entitlement and tabulated as shown in Figure 5-11 (a). Then the allocation volume expressed as percent of total entitlement is plotted against the concerned announcement date as shown in Figure 5-11 (b). As the allocation volume remain constant between two consecutive allocation announcement dates, the graph will be a step line graph which can be defined as Allocation Curve for the year 2013-14.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-11: Preparation of Allocation Curve.

( (A) All cumulative allocation volume are converted into % of entitlment and (B) plotted to get alloction curve (B))

5.5.4.Step-2: Measurement of area under the curves

The allocation curve will have a lowest point on the date when allocation is lowest and highest point on the date when allocation volume is highest. The area under the allocation curve between initial (start) and end of allocation announcement date can be defined as allocation curve area. Likewise, considering the highest allocation as allocation for whole year, we will have a straight line from the initial allocation announcement date to the ending allocation date which will be parallel to the date axis. We can define that curve as highest allocation curve and the area under the curve can be defined as highest allocation curve area.

Abbildung in dieser Leseprobe nicht enthalten

Example:

As shown in Figure 5-12, in the year 2013-14, the allocation curve have the lowest point at the initial allocation date (1st July, 2013) and highest point at the ending allocation date (15th April, 2014) which can be defined as the lowest (36%) and highest (81%) allocation respectively. On the other hand, considering the highest allocation (81%) as allocation for whole year, we will have a straight line from the initial allocation announcement date to the ending allocation date which will be parallel to the date axis (X- axis).

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-12: Preparation of area under curves.

( F i gure (A) is showing the area under allocation curve and (B) is

showing the area under highest allocation curve)

5.5.5.Step-3 : Measurement of Consistency

The consistency can be measured as the ratio between allocation curve area and highest allocation curve area and define as Consistency Index (CI). Higher the value of consistency index, more the consistency in decision making. We can express the CI as % also.

Abbildung in dieser Leseprobe nicht enthalten

Based on the value of Consistency Index (CI), We classify the allocation to GS decision making process in three band as low, moderate and good as follows.

Abbildung in dieser Leseprobe nicht enthalten

In the low consistency band, the allocation decision may observe very low allocation in the initial stage that will continue up to a certain date and before the ending there may be some rise in allocation which may be 2 or 3 times more than the initial allocation. In the moderate band, allocation may observe better allocation than the lower band initially which will observe low growth in regular interval until the ending allocation. In the high consistency band, allocation line will be nearly straight line with minimum vertical movement. However, allocation volume may vary from too low to too high but remain almost constant which is most desirable.

Example:

For the year 2013-14, consistency in general security allocation process can be expressed as,

Abbildung in dieser Leseprobe nicht enthalten

So, in the year 2013-14, consistency of GS allocation decision making process was 76% which is in the moderate band. The allocation curve start from 36% and finally reach to 81% with continuous upward movement in regular interval.

5.5.6.Hypothesis judgement

As per hypothesis of best allocation scenario, if allocation remain constant all over the year,

Abbildung in dieser Leseprobe nicht enthalten

On the other hand, for worst case when allocation curve suddenly rise to 100% from 0 %,

Abbildung in dieser Leseprobe nicht enthalten

So, we can say that the framework acknowledge the hypothesis.

5.6 Application of additional hydrological information inwater allocation

One of our research objective is to investigate the usefulness of additional hydrological information in the water allocation process. We have found that expected natural inflows into the reservoirs are taken as long term historical minimum annual inflow while calculating the total accessible water at the beginning of the each water year in the Murrumbidgee catchment. So, we focuses on to replace this historical minimum inflow values by the inflow data simulated by global hydrological model as additional information while other data remain same in our allocation model. We have found that Water Cycle Integrator (WCI) portal developed under EartH2Observe (E2O) provide public access to water Resources Re-analysis (WRR) and remotely sensed datasets of various hydrological and meteorological variables such as total runoff, precipitation, soil moisture etc. These dataset consists of output of ten different global models that are divided into two types. WRR1 has coarser spatial resolution of 0.5o x 0.5o (approximately 50 km at the equator) for a duration of 1979 to 2012. However, WRR2 has finer spatial resolution and longer period of data. We found that water Resources Re-analysis: tier-2 (WRR2) total runoff dataset simulated by three land surface and one global hydrological model are available in WCI portal. All of these models were forced using the Multi-Scale Weighted-Ensemble Precipitation (MSWEP) dataset. Land surface models (LSMs) usually works in more physically based way and solve both water and energy balances whereas Global hydrological models (GHMs) works in more conceptual way and solve only the water balance to represent hydrological processes (Beck, 2017). An overview of the dataset and models are listed in Table 5-2.

Table 5-2: Overview of Global Models

Abbildung in dieser Leseprobe nicht enthalten

[Source: Schellekens et al. 2017; Dutra et al., 2015]

These total runoff data sets has 0.25o x 0.25o spatial (approximately 25 km at the equator) and monthly temporal resolution for a period starting from 1980 to 2014. Simulated total runoff are available in evenly spaced grid cells. In accessing the data from WCI portal, we prepared the shapefile of the catchment of both reservoirs using SRTM 90m DEM in Q-GIS as shown in Figure 5-13. To extract the monthly total runoff for two catchments, we upload the shapefiles in WCI portal (https://wci.earth2observe.eu/) and downloaded the runoff data for each model as .csv file. The observed data required to run our allocation model is available for 2011 to 2016. Hence, the usefulness of additional data can be investigated for the year 2011-12, 2012- 13 and 2013-14.

Abbildung in dieser Leseprobe nicht enthalten

Figure 5-13: Catchment shapefile of the Blowering and Burrinjuck reservior.

We have annual observed natural inflow into the reservoirs and the conservative predicted natural inflow that has been used at the start of each year in making initial AWD for these years. We define the observed annual inflow volume as the perfect information as this is the volume of water that entered into the reservoirs in the selected years. So we simulate the allocation model with this observed annual inflow data considering decision makers would have this information at the beginning of the year and used it in making AWD. Form the simulated GS water allocation, we measure the consistency index (CI) and define it as CI(Perfect). Moreover, we have simulated the allocation model with the conservative natural inflow (also in emulating observed GS allocation) and measure the consistency of the GS allocation process and define it as CI(Conservative). We also simulate the allocation model with each data set and arithmetic mean of four datasets (ensemble mean) of the global model calculated by equation (5-12) and the consistency index of the GS allocation was measured for each scenario and define it as CI(Addl Info). Inflow data that has been used in the allocation model are presented in the Table 5-3.

Abbildung in dieser Leseprobe nicht enthalten

Table 5-3: Inflow data for different simulation scenario

Abbildung in dieser Leseprobe nicht enthalten

Consistency index of GS allocation process computed for each global data sets are evaluated against the consistency index of conservative and perfect inflow condition following the usual form of skill scores (Stanski, 1989):

Abbildung in dieser Leseprobe nicht enthalten

The Relative consistency index (RCI) is therefore a score between 0 and 1, which is relative to the perfect consistency (CI=1) and worst consistency (CI=0). If RCI value become +ve, it implies consistency increases due to application of additional information. In contrast, if it became zero (0), the use of additional information will not give any benefit than the existing condition. However, if the value become –ve, it implies the additional information will decrease the consistency of the decision making process than observed.

CHAPTER 6

Results

T his chapter consists of the results obtained in this research work in connection to the objectives set and in answering the research questions. It includes detailed calculation of water allocation for the year 2012-13, comparison between different year’s initial allocation and the reason behind the revision of General Security (GS) allocation. An analysis of water allocation to GS with respect to the historical inflow conditions is also provided. Moreover, the model results in emulating the historical allocation and their consistency is included. Assessment of additional information in the decision making process is also explained.

6.1 Available Water Determination (AWD)(2012-13)

The Available Water Determination (AWD) process in the Murrumbidgee region uses several hydro-meteorological observed and predicted data to compute the total available water, and then follows a set of established rules to allocate the available water to various categories. This process is almost similar in every year as described in section 4.3. At the beginning of each year, to make the initial AWD, a detail calculation is done, which gives insight into the allocation process. Initial AWD for the year 2012-13 is explained here.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-1: Determination of water that will be available in the year 2012-13.

The process of calculation of total volume of accessible water for the year 2012-13 is shown in Figure 6 1. First of all total water stored in the storages are measured which was 2,533 GL. Expected inflow into the storages for the next 12 months are calculated as 1,204 GL and total water that would be available in the system was measured as 3,737 GL. Annual losses from the system and essential reserves was calculated as 970 GL which is deducted from 3,737 GL and finally the water that could be allocated to the licensed users were calculated as 2,767 GL.

6.1.1.Storage

As shown in Figure 6-1, the total storage reserves (2,533 GL) at the beginning of the 2012-13 water year are calculated from the total amount of water stored in Blowering dam (1,551 GL), Burrinjuck dam (966 GL), Berembed Weir (1.3 GL), Gogeldrie Weir (0.8GL), Hay Weir (2.9 GL), Redbank Weir (5.5GL), Maude Weir (6.2GL) and Tom Bullen Storage(0.5GL).

6.1.2.Expected Inflow

Expected inflow (1,204 GL) measurement is derived mainly based on the statistical long-term annual minimal natural inflows (233GL) into the storages, and the release to Blowering dam from the Snowy–Tumut Development that is operated by Snowy Hydro Limited. Their operation falls under the Snowy Water License, which indicates a minimum of 1,026 GL of water is required to be released into the Blowering dam annually. This is known as “Required Annual Release (RAR)” into the Murrumbidgee catchment. However, if the inflow into the snowy scheme is found to be reduced when compared to the historical inflow due to drought condition, the risk of drying out of snowy storages arises if RAR has to be ensured. To overcome this situation, a reduction to the RAR is applied by Dry In ow Sequence Volume (DISV) reduction (SHFS, 2016). Moreover, there is no restriction of timing and volume of releases from snowy scheme to the Blowering dam. Hence, sometimes the releases become more than the RAR for a given year which is adjusted by deducting the extra volume in excess of RAR (known as pre-release) from the coming year’s RAR. In addition, though the water year spans from July – Jun, Snowy Hydro Limited measure the releases from May to April. Thus, some water from RAR is usually released before start of the water year (April to June). Provisions known as Relaxation Volume, Reduction Agreement Payback, Snowy license Shortfall Payback etc also play role in determining the expected inflow from Snowy–Tumut Development into the Blowering dam. After considering these factors, expected inflow from Snowy–Tumut Development to the Blowering dam was calculated as 837 GL. Inflow that was on the way due to recent rainfall events and future usable tributaries inflow into the downstream of the two major reservoirs are also included (134GL) in future inflow measurement. It is to be noted that the components of expected inflow measurement vary significantly from year to year depending on the climatic conditions in the catchment.

6.1.3.Reserve and Losses

Detailed calculation of losses that may occur in the next 12 months have been calculated. The main loss is expected to come from the transmission losses (526 GL) that may occur naturally (i.e seepage, evaporation, operational losses etc.) to deliver water from the storages to the various users spread over the catchment. Water that is stored in the major storages and downstream weirs which cannot be extracted under normal operating conditions is known as dead storages (29.6 GL). Moreover, 50 GL of water is taken as minimum reserve for both dams which is also excluded. The 60 GL of water is calculated as post February inflow into the storages, which will not be available during the cropping season, is also excluded. Evaporation losses from the storages for the whole year are taken as 86 GL based on the climatic conditions and the forecast. A minimum daily flow at the end of the river (Balranald) has to be ensured to maintain the connectivity from the Blowering and Burrinjuck to the downstream river for the whole year. This is known as End of system flow. This is calculated as 218GL, considering provisions in the Water Sharing plan (WSP, 2003). These volumes vary from year to year depending on the climatic conditions also.

6.1.4.Allocation to the general Security

The accessible water (2,767 GL) is the final volume of water left for allocation to the licensed users. As described in CHAPTER 4, allocation is done by following the priorities defined in water sharing plan. A simplified process of allocation is shown in Figure 6-2. The calculation begin with the accessible water (2,767 GL) which is shown at the top of column (B) as balance water. Carryover got preference over all licensed users and these requirements are calculated based on previous year’s unused water. After allocating required volume of water for Carryover of GS (511 GL) as shown in column (A), new balance water (B) is calculated as 2,256 GL(2,767 GL - 511 GL).

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-2 Starting (1stJuly, 2012) AWD to various categories of WAL.

(Data source: water.nsw)

In this way, after deducting the allocation to a license category (column A) from balance water (column B), the new balance water is calculated. After providing water to high priority categories following the rules specified in Water Sharing plan, remaining water is given as new allocation to GS (1,211) which is added with the carryover (GS) to get total allocation for GS as 1722 GL (91% of total entitlement).

6.2 Comparison of Initial AWD in different years

The detail calculation process as described in section 8.1 is usually followed to announce initial AWD in every year. A comparison of the observed initial AWD that was calculated for the years 2010-11 to 2016-17 is shown in Figure 6-3. Total accessible water for AWD is calculated by adding storage volume (sky blue bar) with expected inflow (deep blue bar) and deducting reserves and expected losses (red dotted bar). It is evident that the accessible water varies a lot from year to year. Storages volume can be measured physically but the inflow and losses are uncertain as they are predicted value. The prediction of losses were almost same for each year but the predicted inflow volume varied year to year especially in 2012 and 2015. The reason behind this is the expected inflow from the snowy scheme to the Blowering.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-3 Initial AWD for the year 2010-11 to 2016-17

(Data source: water.nsw)

The year 2010 was still part of millennium drought. As a result, expected inflow into the storages are taken more conservatively than usual, which provides low computed accessible water. By the end of 2010-11, the drought period ended with heavy showers in the catchment which contribute huge quantities of inflow to the reservoir by the end of June, 2011. Therefore, the computed quantity of accessible water was also high at the start of 2011-12. This trend was also found in 2012-13 when the initial GS allocation was highest among the previous years. However, due to the low volume of water in the storages at the end of June, 2013, accessible water as well as GS allocation decreased considerably and kept within a narrow range for consecutive years. A good co-relation between accessible water and allocation to GS is also visible in the graph due to the fact that high priority needs are always meet first which are almost same in every year. So, the difference between accessible water and GS allocation is almost same in every year.

6.3 Revision of the allocation decision:

The reason behind the revision of GS allocation was found to be the increase in inflow volume into the reservoirs, as compared to the original expected inflow. All other categories get almost full of their entitlement at the beginning of every year, thus allocation to GS improves as the inflow increases until full entitlement is reached. As is shown in Figure 6-4, actual observed inflow into the reservoirs is always (much) higher than the expected inflow, as the expected inflow is based on a minimum historical inflow. As the year progresses, the inflow into the reservoirs typically exceeds the expected inflow volume for that period. As a result, this extra volume of water then becomes available for allocation, and the GS allocation is incremented with the increase in inflow condition. For instance (Figure 6 4), in the year 2013-14, the total expected annual inflow was 718 GL and the initial GS allocation was 680 GL. However, as the inflow scenario improved by the course of the year, allocation to GS was also increased and finally reached at 1,532 GL by the end of the season whereas the observed inflow also reached to 2,094 GL.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-4: Comparison between observed and expected inflow.

(Data source: water.nsw)

6.4 Historical water allocation to GS

Allocation to GS depends on the availability of water and natural inflow conditions in the Murrumbidgee valley. The long term average annual natural inflow into the Murrumbidgee system (inflows to the dams and the downstream tributaries) is 3,061 GL (NSW, 2010). However, from late 1996 to mid of 2010, the valley was affected by the Millennium Drought when the rainfall, as well as inflow into the reservoirs were significantly below normal. As shown in Figure 6-5, severe shortages in inflow as well as in water availability were observed during 2006-07 to 2009-10. For instance, the inflow fell to 246 GL in 2006-07, which was well below the long-term historical inflow into the valley. On the other hand, due to dominance of La Niña conditions in 2010-11, 2011-12 and 2016-17, heavy rainfall around the region resulted in excess inflow into the system.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-5: Comparison of actual natural inflows with average inflow.

(Data Source : water.nsw)

By comparing actual annual inflow with the long term average inflows, years are divided into wet, dry and very dry years. As illustrated in the Figure 6-6, those years where the inflow is below one-third of the average inflow are defined as very dry (2006-07, 2007-08, 2008-09 & 2009-10) and those above this threshold are defined as dry year (2004-05, 2005-06, 2012-13, 2013-14, 2014-15& 2015-16). For some years, observed inflows are significantly higher than average. These are defined as wet years (2010-11, 2011-12 & 2016-17). During the period from 2004 to 2016, no year had inflow close to the average inflow, implying that there are no normal years.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-6: Total annual natural inflows into the Murrumbidgee valley.

(Data Source : water.nsw)

6.4.1.Water allocation to GS in very dry years

Water allocation to GS is usually very low in the very dry years. As can be seen in the Figure 6-7, in the year 2007-08, the initial allocation was as low as 7.5%, and final allocation was 20.5% of the total entitlement, which is regarded as minimum allocation in the history of the Murrumbidgee. At the start of this very dry sequence in July, 2006, two major storages (Burrinjuck and Blowering) were holding 1,219 GL of water and allocation to GS was 554 GL (27% of the entitlement). Due to historically low inflow during this year, allocation was increased by only 1% and thus reached to only 574 GL by the end of the season. Due to lack of new incoming water into the system, the majority of demand was met from reservoir storage which consequently decreased the storages volume to as low as 461 GL (17% of total storage capacity) by March (end of peak irrigation season) finally reaching to 741 GL by the end of June, 2007 due to decreased demand and improved inflow condition. Moreover, carry-over was also decreased to as low as 50 GL, which means almost all water was used. During the year 2007-08, initial allocation was very conservative and allocation increased from 151 GL (7.5%) to 415 GL (20.5%) by the end of season. However, the carryover to the next year (2008-09) was 205 GL which was almost half of the allocated water and means usage was also low. During two consecutive years 2008-09 and 2009-10, initial AWD did not allocate any water except the carryover volume from the previous year. In both years, revision in allocation decision was done several time and final allocation was below 40% of total entitlement. Due to conservative use of water, reservoir storage did not drop below 900 GL after 2007 though inflow was very low throughout the years. It is evident that carryover acts as a risk management tool for GS license holders in these very dry years and provides insurance of some amount of water at the start of every year to continue farming. In these periods, high security users are also allowed to carryover water though the water sharing plan officially does not allow it. Due to severe water shortage, water sharing plan for the Murrumbidgee was suspended on 10/11/06 and recommenced again after the dry period broke on 16/09/2011.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-7: Historical water allocation for GS in very dry years.

(Data Source : water.nsw)

6.4.2.Water allocation to GS in dry years

Among the years defined as dry, 2004-05 & 2005-06 experienced the Millennium Drought and 2012-13 was followed by consecutive wet years. Therefore, allocation in those years shows a different trend to the other years (Figure 6-8). In 2004-05 and 2005-06, initial stored water in major two reservoirs was almost 20% of storage capacity (494 GL & 545 GL) and the starting allocation was 17% of the total entitlement. However, improved inflow scenarios in 2005-06 offer 10% higher final allocation than 2004-05. In both years, there were many revisions of the allocation decisions (10-12 times) and carryover volumes were comparatively lower than other years. In contrast, with 91% of initial allocation, 2012-13 observed the highest allocation in GS before major planting season (October) and 100% at the end of November. This higher allocation was made possible as the two major reservoirs were jointly holding 94% (2,508 GL) of their storage capacity and the trends of inflow into the reservoir from previous two consecutive years provided abundant of water in the region. Due to higher allocation before the planting season, the use of water was also highest in this year and carryover to next year was also comparatively low (18%) (Figure 6-8). In other years from 2013-14 to 2015-16, initial allocation ranges between 31-36%, resulting from moderate storage of water in reservoirs. Though final allocation in 2013-14 and 2014-15 observed almost a 45% increase from the initial level and reached above 80%, the final allocation for 2015-16 was nearly 60% due to lower inflow than other two years. Contribution of carryover from previous years in the initial allocation and number of revision in allocation decision were also higher in these years.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-8: Historical water allocation decision for GS in dry years.

(Data Source : water.nsw)

6.4.3.Water allocation to GS in wet years

During the all wet years, final GS allocation reached the full entitlement volume (Figure 6-9). During the transition between the very dry and wet season on 2010-11, the starting allocation was equal to the carryover volume (480 GL, 25% of entitlement) which resembles the previous dry years of 2008-09 and 2009-10. However, due to significant increase of inflow into the system, allocation crossed 75% before end of planting season around mid-October and reached 100% by mid of December. This means an almost 75% increase in allocation within six months. Change in allocation volume is also observed, with significant improvement in a single month, in September 2010 when within 15 days (1st to 15th September) 680 GL (36%) of extra water has been allocated to GS. Moreover, as a result of this significant improvement in available water, storages volume rose from 1,181 GL to 2482 GL by the end of June, 2011. Therefore, the starting allocation in 2011-12 was as high as 71% of total entitlement including 26.5% of carryover from the year ending at June, 2011. During this year, allocation increases gradually with the improvement of inflow and by the end of November reached to 100% much earlier than previous year. Surprisingly, from 2010-11 to 2012-13, three consecutive years experienced full allocation to GS after prolonged drought period. After four consecutive dry years, 2016-17 started with higher inflow sequence which made possible to allocate 40% of total entitlement to GS including 19.6% of water from carryover. In this year, by the mid of November, 100% allocation was given to GS after 8 revisions in allocation decision.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-9: Historical water allocation decision for GS in Wet years. (Data Source : water.nsw)

6.4.4.Summary of the analysis

It is evident form this analysis that the allocation process is well monitored all through the year and never go downward even in the historical driest year, 2006-07. It is good for the farmers as they don’t have to face crop damage in the middle of the year. If it goes downward, the irrigator will find that they have no water to deliver into the field at the middle of the season though they have planted according to their initial allocation. It indicates the robustness of the GS allocation decision making process and become possible due to the conservative approach in estimating the available water. Though it shows good result for dry years but causes frequent revisions in allocation during wet years. Therefore, wet years experiences inconsistencies in water allocation though there was no shortage of water in reality. It hampers to extract optimal benefit from the available water resources. However, it may be overcome by using additional information or seasonal prediction tools in decision making process which may provide better information regarding future inflow to take decision in an appropriate way. As a consequence, it will help to provide a reliable outlook about the end of season allocation to the irrigators as well. Thus, irrigators can plan and crop more with a clear vision regarding the quantity of water they may get up to end of season.

6.5 Emulation of water allocation process to GS

Though initial allocation decisions mostly depend on the available water and conservative inflow calculation, revisions to the allocation is complex, and depends on various factors like quantity and intensity of rainfall, catchment conditions that affects rainfall-runoff, operational requirements etc. (DPI Water, 2015). The model that has been described in section 5.4 has been used to emulate the revision in GS allocation for the year 2011-12 to 2016-17 and compared with the actual observed GS allocation. In doing this, observed actual starting reservoirs storage for each year have been taken as initial reservoirs storage in the model. The initial expected inflow, losses and other reserve values for each year are taken from the General Purpose Water Accounting Report (GPWAR) of NSW and shown in Appendix D and kept same all around the year.

The simulated GS allocation has been plotted with the actual observed allocation and shown in Figure 6-10. Simulated allocation was evaluated with three performance metrics Pearson’s correlation coefficient (r), Percent bias (PBias), Kling–Gupta efficiency (KGE) as described in section 5.4.5. The correlation (r) value ranges from 0.84 to 0.99 which indicates good linear relationship with the observed value. PBias values ranges from -11.89% to 6.18%, which indicates the model sometimes underestimates and sometimes overestimate than the observed value. KGE value ranges from 0.43 to 0.93 which indicates better fit of simulated value with the observed.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-10: Actual observed and modelled GS allocation (2011-2016). ( P e arson’s correlation coefficient (r), Percent bias (PBias), Kling– Gupta efficiency (KGE) values for the model output fall within the acceptable range)

During a season, concern authority consider catchment condition, usage, losses and other factors which may influence in the revision of decisions. Therefore, the modelled GS allocation sometimes shows better allocation than the actual and sometime reduced.

6.6 Allocation and Release

While emulating the observed allocation decision of GS with the model, the initial storage in the reservoir at the start of the water year has been taken same as the observed volume. However, after simulation, it was found that at the end of the year, the modelled storage become lower than the observed reservoir storage (Figure 6-12) though the allocation volume were almost similar. This may have happened due to the fact that nearly all allocated water in the model is being released form the reservoir. In reality, a portion of allocated water is forfeited due to the fact that the users don’t use all of their allocated water and “bank” some water to be carried over to the next year. A maximum 30% of the entitlement can be carried over to the next year by GS and conveyance categories. Figure 6-11 (A) shows the historical data (2009- 10 to 2016-17) regarding volume of water that has been allocated to the license categories but remain unused at the end of the year. The Blue bar represents the forfeit volume and the Red bar represents water that has been carried over to the next year. As can be seen in the Figure 6-11 (a), in the year 2010-11 and 2011-12, the volume of unused and forfeited water was very high.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-11: Unused water and Carryover

(Data source: GPWAR, NSW)

To account for this, a factor was introduced in the allocation model to decrease the released volume compared to the allocated volume. This would be expected to provide a better fit of modelled to the observed storages. It was, however, found that for different year’s reservoir observed and modelled storage fitted good for different values of this release percentage. By trial and error, it was found that if 78% of the total allocated water is released from the reservoir, modelled end of year storage and carryover volumes are close to the observed volume for all years. Figure 6-11 (b) shows the comparison between observed and the modelled carryover volume when model release is 78% of the total allocated water. On the other hand, Figure 6-12 is showing comparison between modelled and observed storage in the reservoirs. As can be seen, for the year 2014-15, modelled PBias has been improved form -9.55 % to 5.84 % and KGE value also shows better fit by increasing from 0.76 to 0.86 due to the application of release factor. Similar improvement happen for the year 2015-16 where PBias improved form -20.28% to -6.32% and KGE improved from 0.44 to 0.92. It should be noted that carry over volume is calculated by deducting the total end of year release from the total allocation volume against the GS and Conveyance category in the model.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-12: Reservoir storage for different release-allocation ratio. [ P e arson’s correlation coefficient (r), Percent bias (PBias), Kling– Gupta efficiency (KGE)]

6.7 Consistency in GS allocation in previous years

Based on the water availability and inflow conditions in the Murrumbidgee catchment, the years that were evaluated have been divided into very dry, dry and wet as described in section 0. The framework that has been described in section 5.5, has been applied to measure the consistency of the GS allocation decision making from the year 2004-05 to 2016-17. As shown in Table 6-1, the consistency index varies from 0.44 to 0.96. The index values shows that, consistency was good in the dry years as all the years fall within moderate to high consistency band. This may due to the fact that in dry years the inflow condition does not change so much compared to the conservatively estimated expected inflow. As a result, the allocation decisions become more consistent. On the other hand, in very dry and in wet years, the consistency index varies from high to low. In wet years, the allocation typically starts from a low value and due to ample rainfall, allocation rises up to 100% with several revision. Conversely, in dry years the GS allocation starts with very low value, but by the end of the cropping season, the allocation value suddenly rises to some extent. As a result, though the increase in allocation is low but due to the timing of the allocation and the ratio of increase with respect to the initial allocation being very high the consistency index is low. The increase in allocation in that time has no value to the farmers expect carry over the water to the next year.

Table 6-1: Consistency of GS allocation decision making (2004 to 2016)

Abbildung in dieser Leseprobe nicht enthalten

6.8 Model simulation with different inflow data

As described in section 5.6, we extracted total runoff data simulated by four global model. One of our research objective is also to investigate the usefulness of this dataset in water allocation process in the Murrumbidgee region. It should be noted that the output of this dataset in our allocation model don’t have any affect to the actual or official allocation process rather to inform the irrigators about the likely volume of water they are going to have in the coming irrigation season. With better information regarding probable end of season allocation, irrigators can plan their cropping choice rationally to maximize their farm returns. It will also reduce the dependency on buying expensive water from the water market at the end of the season to finish the harvest when expected water did not come into the allocation announcement. Therefore, the output of this analysis can be treated as an allocation advisor statement to the irrigators.

We simulated our allocation model with each dataset as described in section 5.6. Result of each simulation is described below.

6.8.1.Simulation result using perfect information

As shown in Figure 6-13, Simulation result using perfect information shows that the allocation will be 100% for the year 2011-12 & 12-13. The consistency index (CI) gives a value of 1 which indicates perfect allocation decision. However, in the year 2013-14, allocation improves from 60 % to 69% by the mid December. However, with perfect information, it is expected that the allocation will be same from start to end of season without any increase or decrease. But in reality, it may not happen due to the flow variability in a season. The model is calculating resources in each time step where the storage volume changes due to the observed inflow. Conversely, expected inflow is reducing after every time step following the long term average inflow curve shown in Figure 5-6. Therefore, it may happen that the inflow entered into the reservoir before the mid of December is higher than the expected inflow, which provides more water to the model for allocation. Due to the revision in allocation the value of CI is found to be 0.92 which indicates slight variation in decision making process. A fall in reservoir storage was observed from the simulation which dropped to almost 500 GL by the mid of April.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-13: Simulation result using perfect inflow data (2011-14)

6.8.2.Simulation result using conservative inflow

Simulation result using conservative inflow shows the decision making process was better in the year 2012-13 when the allocation rises from 91% to 95% by the end of November (Figure 6-14). On the other hand, in the year 2013-14, allocation decision revised a lot and increased from 36% to 85% by the mid of April. However, major improvement happened until mid of November and afterwards change was low. The CI value reflects this variation and gives a value of 0.76 for 2013-14 whereas 2011-12 and 2012-13 gives a value of 0.90 and 0.95 respectively. Storage in the reservoir dropped close to 1,000 GL by May 2013 but recovered by September 2014 and again show a declining trends and goes below 1,000 GL by end of March, 2014.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-14: Simulation result using conservative inflow data (2011-14)

6.8.3.Simulation result using inflow data of W3RA global model

Simulation result (Figure 6-15) using inflow data of the W3RA global model shows perfect consistency with 100% allocation to GS for the year 2011-12 and 2012-13. However, In the year 2013-14, allocation shows a decreasing trend and fall from 100% to 78% in July 2013. Afterwards, the allocation does not change and the CI value gives 0.87. Storage in the reservoir went below 500 GL by the end of March, 2014 but afterwards shows increasing trends.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-15: Simulation result using inflow data of W3RA model (2011-14)

6.8.4.Simulation result using inflow data of HTESSEL global model

As shown in Figure 6-16, 100% allocation to GS was achieved using HTESSEL global model inflow data which give CI value 1 for the year 2011-12 and 12-13. However, in the year 2013- 14, allocation rises from 59% to 69% by the end of December and value of CI was found 0.88. Storage in reservoir fall close to 500 GL by the end of March, 2014 and shows an increasing trends afterwards.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-16:Simulation result using inflow data of HTESSEL model(2011-14)

6.8.5.Simulation result using inflow data of SURFEX-TRIP global model

As shown in Figure 6-17, 100% allocation to GS was achieved using SURFEX-TRIP global model inflow data which give CI value 1 for the year 2011-12 and 12-13. However, in the year 2013-14, allocation rises from 46% to 66% by the end of December and value of CI was found 0.87. Storage in reservoir fall close to 600 GL by the end of March, 2014 and shows an increasing trends afterwards.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-17: Simulation result using inflow data of SURFEX-TRIP model

6.8.6.Simulation result using inflow data of PCR-GLOBWB global model

Simulation result (Figure 6-18) using inflow data of PCR-GLOBWB global model shows perfect consistency with 100% allocation to GS for the year 2011-12 and 2012-13. However, In the year 2013-14, allocation shows improving trend and rises form from 40% to 65 and the CI value was calculated as 0.84. Storage in the reservoir went below 700 GL by the end of March, 2014 but afterwards shown increasing trends.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-18: Simulation result using inflow data of PCR-GLOBWB model (2011-14)

6.8.7.Simulation result using mean inflow of global model datasets (Mean-GHM)

As shown in Figure 6-19, 100% allocation to GS was achieved using mean of all global model inflow data which give CI value 1 for the year 2011-12 and 12-13. However, in the year 2013-14, allocation rises from 57% to 68% by the end of December and value of CI was found 0.88. Storage in reservoir fall close to 500 GL by the end of March, 2014 and shows an increasing trends afterwards.

Abbildung in dieser Leseprobe nicht enthalten

Figure 6-19: Simulation result using mean inflow of global model datasets (Mean-GHM)

6.9 Relative consistency index (RCI)

The relative consistency index is calculated using equation (5-13). The result is shown in Table 6-2. It is observed that every global model data simulation give RCI values of 1 for the years 2011-12 and 2012-13, which is the perfect value. However, for the year 2013-14, RCI value ranges from 0.50 to 0.75. However, As the values are positive, additional inflow data from the global models give sign of improvement form the existing condition. Inflow data from HTESSEL and mean of four global data sets (Mean-GHM) provide better consistency.

Table 6-2: RCI value for different global inflow data:

Abbildung in dieser Leseprobe nicht enthalten

It is also observed that the bias in global model dataset has influence in the allocation model result. In the two wet years, 2011-12 and 2012-13, when the reservoir storage as well as inflow was significantly high, the effect of bias of data sets were not visible in the allocation. However, it has a consequences in the 2013-14, when the reservoir storage was low as well as the inflows. In the year 2011-12, global data set has negative bias (underestimate the inflow), but as the reservoir storage was high, the GS allocation reached to 100%. It happens due to the fact that the allocation model calculates the resources based on expected inflow (global model data) and reservoir storage which provide higher value than the cumulative entitlement of all license categories. So, form the beginning of the year, the resources become surplus than the demand. Therefore all license categories get 100% of their entitlement. In 2012-13, W3RA, HTESSEL and SURFEX-TRIP has positive bias (overestimate the inflow) and PCR_GLOBWB has negative bias. Yet, none of them are visible due to high storage in the reservoirs and inflow condition. However, the influence of bias in global dataset become visible in the year 2013-14. In that time, W3RA model data has higher positive bias. As a result allocation model estimate higher volume of resources at the start of the year by adding reservoir storage and expected inflow (W3RA data) and provide full allocation to GS. But within one month, allocation model observe less inflow than the overestimated expected inflow, which drives the model to decrease the resources. As a consequence, being lower priority category, a fall in allocation volume to GS is observed. This is very alarming for agriculture, especially for annual crops. Annual crops are very susceptible to the availability of water for irrigation. Usually farmers plant crops with respect to the allocation information. Therefore, with information of higher allocation, they will plant more. However, if at the middle of the year they are faced with curtailments of water volume, either they have to buy required water from the water market at a high price or leave the crops without irrigation. Both will result in financial losses to the irrigators. Besides, due to higher allocation at the beginning of the year, the release was also higher which decreases the reservoir’s storage that results in dropping the reservoir storage within that time. On the other hand, in that year SURFEX-TRIP, PCR_GLOBWB has a negative bias, so the resources increases as the year progress and the allocation process observes more revisions. The HTESSEL model data set has lower bias. As a consequence, less revisions are observed with respect to the simulation with perfect information. However, it is found that the mean of global data sets provide lower bias in all years, as well as better consistency. Besides, it is robust to choose the mean of all datasets when the model outputs show large variation. Moreover, it make sense to take mean when several model output is available for use.

CHAPTER 7

Discussion and Conclusion

T his chapter contains brief discussion regarding this research work, summarizes the research findings and its relevance to the objectives. It also includes the limitations of this research and prospect of further research.

7.1 Discussion

This research reveals that the annual allocation process in the region is dependent on the availability of water and natural inflow conditions. In normal and wet years, when catchment inflow exceeds the long-term average inflow volume, allocation to General Security (GS) category (for irrigating annual crops) usually reaches the full entitlement after several revisions. On the other hand, in dry years, the allocation starts low and fewer revisions are observed. At the start of every year, the decision-makers use a historical minimum annual inflow (233 GL) into the two upstream reservoirs as the future expected inflow to the reservoir along with other sources of water in calculating the total volume of water that will be available in the current year. However, as we have analyzed in our study period from 2011 to 2017, in most years this volume increases significantly as the year progresses. For instance, in the year 2011-12, the observed inflow was found to be 2,971 GL, which is 12.8 times higher than the original expected inflow of 233 GL. In other years, the increase of inflow was around 5 times higher than the expected value. As a result, the volume of water available for allocation increases as the year progresses. From these observations, it becomes clear that there is an inconsistency in the allocation decision to GS water as has been observed in previous years and that this is mainly due to the conservative prediction in future inflow.

The allocation model that was developed in this research uses several hydrological observed and predicted data to compute the availability of water and finally to allocate the resources between different categories of user based on the established rules that have been established in the region. It also simulates the reservoir operation based on the demand and the volume of water allocated to different categories. Hence, it can track the reservoir storage at every time step (chosen as one day), including provisions for spill as well as minimum storage level indication, which can give a better idea to the decision makers to gain insight into the effect of decisions they are going to take. This model can be used to test any dataset to investigate its effect on the allocation decision.

After simulating the allocation model, we found that the model was performing well in emulating the previous decision making process. However, though the model simulates release of water from the reservoir based on the allocation volume to different categories and demands, the reservoir storage was found to be decreased more significantly than the observed level though the allocation volumes were almost similar. The reason behind this discrepancy was identified as the unreleased water that is allocated to the users but remained unused by them, and that these users then “banked” as carry-over to the next year. We found that if the release volume from the reservoir is set to a factor of on average 78% of the allocated water, the simulated reservoir level closely tracks the observed levels. Moreover, the carryover volumes we simulate are close to the observed carry-over volumes, thus confirming the results. Therefore, the model developed is found to be useful for simulating the water allocation process, considering the many relevant aspects while allocating water to different categories of users.

To investigate the value of additional data in informing the allocation process, we used reservoir inflow data obtained from four global hydrological model available in the eartH2Observe WCI portal. All other data were kept unchanged as used for emulating historical allocation process to explore the benefit of using additional hydrological data. The benefit of this additional information has been determined by measuring its capacity to improve the course of action in making allocation decisions. The decision making process was simulated for three consecutive years (due to data constraints), spanning from 2011-12. The model result has been evaluated with the framework that has been developed to measure the consistency in the decision-making process. We observed that better consistency in GS water allocation process can indeed be achieved by using additional information. The result shows that in the year 2011-12 and 2012- 13, 100% of the entitlement is allocated to GS at the start of the season, and subsequently not revised. In reality, GS allocation process experienced several revisions in those years. The year 2013-14 shows a slight revision when using the new data but does provide better consistency than the observed. The improvement in the decision making process is also reflected in the consistency index and the relative consistency index (CI) values. It was observed that the consistency index reached to its perfect value, 1, which means the GS allocation process become consistent and remains the same all through the year for the year 2011-12 and 12-13. However, CI value ranges from 0.84 to 0.88 in case of additional information which is better than the observed value (0.76). As a consequence, the relative index value also shows perfect value of 1 for 2011-12 and 12-13 and positive value ranges from 0.5 to 0.75 for 2013-14 which means the consistency would come close to consistency that could be achieved if perfect information was used. Therefore, it is evident that additional information can provide better consistency in allocation process, which is desirable to farmers as they can then better plan their cropping season.

In the year 2013-14, additional information obtained from W3RA global model shows declining trends in water allocation at the middle of the year. It shows that the starting GS allocation was 100% but it dropped to 78% within one month and then remains constants through the year.

This happens due to the high positive bias (overestimation) associated with the W3RA datasets which shows more water is available to allocate at the start of the year. However, as the year progresses, the model calculates lower resources than the initial estimate due to the actual inflow being lower than the expected. Therefore, allocation to GS is also reduced as it is the least priority category to get water and any decrease in the estimated water availability is translated to a decrease in the allocation to GS. Curtailment of the allocation for GS is not preferred as it may cause damage to the crops and subsequent financial losses. However, after the reduction of the allocation to GS, the allocation volume does become stable before the actual start of cropping season (October), which means it may not affect the user so much in the end. It does, however, decrease the reliability of the decision making process which may have consequences to the end user. Hence, it is evident that the bias in global inflow datasets influences the output of allocation model, especially when the reservoir storage as well as the inflow is low (2013-14). A low bias leads to multiple revisions to increase allocation whereas a high bias (+ve) results in curtailments in the allocation, which is undesirable. The ensemble mean of four dataset was found to perform well and provide less bias. Therefore, it is desirable to use mean of global data sets in the case several datasets are available and bias is unknown.

The simulated reservoir storages were also analyzed when using the additional information. It was found that the reservoir levels were dropping lower than the observed levels. This is due to the full entitlement being allocated in consecutive years, and thus more water being used. However, as we observed there is an additional positive consequence in that less water goes to waste. In the observed data, water was spilled from the reservoirs due to the high inflow starting in March, 2012. This did not occur in the model due to the fact that from the start of the year 2011-12, the allocation was 100% and as per demand the release was also higher in the cropping season. Hence, the reservoir level dropped lower towards the end of the cropping season (February) which gives more space in the reservoir to hold the large inflow during the non- irrigating season. This provides extra water in the reservoir for the next year, which gives scope to allocate the full 100% entitlement in 2012-13 as well. However, in the year 2013-14, the reservoir level was falling close to 500 GL or even below which will cause less allocation in the next year as availability of resources will decrease. This may cause a sudden drop in crop production, which may not be acceptable to the farmers.

The use of global data sets in allocation model is assessed to provide an idea about the end of season allocation at the beginning of the year. The output of the model can be used to provide allocation advisory statement to the irrigators. It is not related to official allocation announcement, but is the information that the farmers may take into account in establishing their own plans as a function of the expected water availability. Overall, we found that application of global datasets as additional information can provide increased GS allocation at the beginning of the year, and shows fewer revisions and results in an improved consistency in the allocation process. It can also provide better information regarding the end of season allocation, which can maximize the profit and minimize the production risk due to the fact that this information is the key factor for making cropping decision (type of crops, area of harvest etc.). Clearly hydrological simulations of expected flow for the forthcoming season are not routinely available. However, seasonal climatic forecasts of rainfall can be used to develop a streamflow forecast in the region. In doing this, locally calibrated model can be a good option which can provide forecast of inflow into the reservoirs more frequently that can be used in our allocation model to make the allocation advisor statement more robust and reliable.

7.2 Conclusion

This research is focused on assessing the consistency in water allocation decision making process, based on the case study of the Murrumbidgee regulated river and assess whether this can be improved by additional hydrological information.

In doing this, we studied the region and process of water allocation from 2011 to 2016, for which detailed information is available, to gain insight into how decisions were taken in previous years. Based on this understanding, a water allocation model has been developed that is capable of emulating the allocation process that has been observed in previous years. Moreover, a framework is developed to assess the consistency of the water allocation decisions, both in their current form, as well as when using additional hydrological information. Finally, we investigate the benefit of informing the allocation process with additional hydrological information to improve the consistency of the allocation decision.

The water allocation process in the Murrumbidgee basin is well monitored and supported by various hydro-metrological data. Even though agriculture is the largest user of water in this region, annual crops such as rice, wheat, corn, tomatoes, cotton are dependent on the water allocation to a General Security license category, which is established at the start of each water year depending on actual and expected availability of water. The allocation process is found to take a conservative view on the expected availability of water through to the end of the season, resulting in lower allocation at the beginning of cropping season. As the year progresses, the allocation is often revised, and tends to increase, leading to an inconsistency between the initial and the final allocation.

The allocation model that has been developed in this research can emulate the water allocation decision making process in the Murrumbidgee region. The model performed well while emulating the GS water allocation process that has been observed from the year 2011 to 2016. The simulated and observed GS allocation for these years has been evaluated using performance metrics such as Pearson’s correlation coefficient (r), Percent bias (PBias), Kling– Gupta efficiency (KGE). The results show that r- value ranges from 0.84 to 0.99 which indicates good linear relationship with the observed value. PBias values ranges from -11.89% to 6.18%, which indicates the model sometimes underestimates and sometimes overestimates the observed value, though these biases are considered to lie in between acceptable ranges. KGE values range from 0.43 to 0.93, which indicates a good fit of simulated value with the observed. The model can generate various outputs such as daily reservoir volume, total resources for allocation, environmental water needs and have provision of spillage and minimum reservoir storage. Moreover, farmers may not use all the allocated water, as they then have a right to use that water in a next year as carry-over. The model is also able to emulate this behavior. Furthermore, it is also possible to investigate the influence on the allocation process with any data sets that provide an expectation of water availability.

The framework developed to evaluate the GS allocation decisions has been applied to assess the past and simulated allocation decision scenario. The framework is unique in nature in that it evaluates the decision process by a value ranging from 0 to 1, which we have defined as a Consistency Index (CI). The allocation decisions is said to be perfect with a CI value of 1, when the allocation remains constant throughout the year. If the allocation changes, the value decreases accordingly. Using this framework, we evaluated the allocation decisions from 2004 to 2016. We found that in general better consistency happens in the dry years as the allocation then starts low and does not improve significantly through the year. This in contrast to normal and wet years, when the allocation starts low and is often revised upwards through the year.

To explore the benefit of additional hydrological information to inform the decision process, we used simulated inflow data from four global hydrological and land-surface models; SURFEX, HTESSEL, W3RA and PCR-GLOBWB, obtained from the EartH2Observe data portal in our allocation model for the year 2011-12 to 2013-14. The model results show that when using the additional data from the global model output, the consistency in the allocation process can be improved. We found that, decision makers could allocated 100% of entitlement to the GS category at the beginning of the relatively wet 2011-12 and 2012-13 water years, which in reality had only been 71% and 91%. Our measurement framework also shows that the consistency index would be 1.0 for those years, whereas in the observed case it was 0.84 and 0.91. For the dryer year, 2013-14, allocation would have been a maximum 69% with a consistency index of 0.88, where the actual observed value was 0.76. Allocation model results shows that the bias in global inflow datasets influenced the output of allocation model especially when the reservoir storage as well as the inflow is low (2013-14). A low bias leads to multiple revisions to increase allocation whereas high bias (+ve) may result in curtailments in the allocation, which is undesirable. The ensemble mean of the four datasets was found to perform well and provide less bias. Therefore, it is desirable to use mean of global data sets in the case several datasets are available and the bias is unknown.

Improvement in consistency implies that it will provide better information regarding the availability of water throughout the irrigation season (end of season allocation) as well as at the beginning of cropping season to the irrigators. Cropping decisions by farmers are dependent on this information, and the more consistent the allocation process, the more production will maximize the benefit from the available water. These findings supports the relevance of using global data in local decision-making processes. There are many areas in the world where ample data are collected and used for making long term planning in water allocation process. However, in the case of the short term or annual water allocation process, decision makers often do not have relevant data or information available due to lack of infrastructure or accessibility. In those areas, the use of global data as additional information can provide support in making more reliable and consistent allocation decision.

7.3 Limitations and assumptions of the study

- In developing the model, we use the expected inflow and losses data as they are used by the NSW water authority at the beginning of each year. Based on this data, we simulate the release for losses and reducing expected inflow based on seasonal variability. However, in the actual case, those data are changed as the year progress based on the climatic condition. Therefore, for better performance of the model, observed midterm data can be used.
- Simulated release from the reservoir is based on the demand curve that we have used in developing the allocation model. However, in reality the demand may decrease due to rainfall over the catchment, or may increase due to the dryness. As a result, model may shows some discrepancies to the actual scenario.
- Global hydrological models are not locally calibrated. So there are uncertainties of using those data in local scale. Therefore, locally calibrated hydrological data can be used which was not available at the time of doing this study.

7.4 Recommendation for future work

- The simulation of the allocation process should be tested for a longer duration by gathering relevant data and more global or local inflow data as additional information to investigate their value in allocation decision making process.
- Development of seasonal inflow forecasting and use of the forecasted inflow into the developed model can be used in the developed allocation model to investigate its usefulness in the allocation process.
- A framework can be built to explore the effect of better consistency of water allocation process on the crop production, as well as on the economy.
- The methodology developed in this research for assessing the consistency in water allocation decision making process can be applied to other catchments where there are similar systems of allocation.
- Water allocation in the early season is of greater value to irrigators than off season allocation. Further research is recommended to improve the consistency measurement considering the timing of allocation and other relevant factors.

Reference

A guide to the Water Sharing Plan for the Murrumbidgee Regulated River Water Source, 2004,DIPNR, NSW.

Aither 2016, Water markets in New South Wales: market outcomes, trends and drivers, AitherPty Ltd.

Asian Development Bank (ADB), 2009, Water rights and water allocation: issues and challenges for Asia. Jeremy Bird, Wouter Lincklaen Arriens, Dennis Von Custodio.

Balsamo, G., Albergel, C., Beljaars, A., Boussetta, S., Brun, E., Cloke, H., Dee, D., Dutra, E., Munõz-Sabater, J., Pappenberger, F., De Rosnay, P., Stockdale, T. and Vitart, F.: ERA-Interim/Land: A global land surface reanalysis data set, Hydrol. Earth Syst. Sci., 19(1), 389–407, doi:10.5194/hess-19-389-2015, 2015.

Beck, H. E., van Dijk, A. I., de Roo, A., Dutra, E., Fink, G., Orth, R., & Schellekens, J. (2017).Global evaluation of runoff from 10 state-of-the-art hydrological models. Hydrology and Earth System Sciences, 21(6), 2881.

Bierkens, M. F. P., & Van Beek, L. (2009). The Global Hydrological Model PCR-GLOBWB.In F. o. E. S. Department of Physical Geography, Utrecht University (Ed.). P.O. Box80115, 3508 TC, Utrecht, the Netherlands.

Burrell M., Moss P., Petrovic J., Ali A., (2017) General Purpose Water Accounting Report2015-2016: Murrumbidgee Catchment, NSW Department of Primary Industries, Sydney

C. Linés, A. Iglesias, L. Garrote, V. Sotés, M. Werner (2018) Assessing the effect of additional information on operational drought management decisions in the Ebro basin.

Critical water planning for the Murrumbidgee Valley, Issue 35, 15 June 2010, NSW office of Water.

de Fraiture, Charlotte; Wichelns, Dennis. 2010. Satisfying future water demands for agriculture.Agricultural Water Management, 97(4):502-511. Special issue with contributions byIWMI authors. doi: http://dx.doi.org/10.1016/j.agwat.2009.08.008

Decharme, B., Alkama, R., Douville, H., Becker, M. and Cazenave, A.: Global Evaluation of the ISBA-TRIP Continental Hydrological System. Part II: Uncertainties in River Routing Simulation Related to Flow Velocity and Groundwater Storage, J. Hydrometeorol., 11(3), 601–617, doi:10.1175/2010JHM1212.1, 2010.

Deliverable No: D.5.1 – Report on the current state-of-the-art Water Resources Reanalysis (WRR1-tier 1), EartH2Observe. (http://earth2observe.eu/files/Public%20Deliverables/ D5.1_Report%20on%20the%20WRR1%20tier1.pdf) accessed on 15March 2018.

Department of Water and Energy (DWE). 2009. Water sharing in the Murrumbidgee Regulated River (Progress report 2004 to 2008)., NSW

Dinar, Ariel; Rosegrant, Mark W.; Meinzen-Dick, Ruth. 1997. Water allocation mechanisms : principles and examples. Policy, Research working paper ; no. WPS 1779. Washington, DC: World Bank.

Doreen Burdack, 2014, Water Management Policies and their Impact on Irrigated Crop Production in the MurrayDarling Basin, Australia

DPI Water (2013), Determining water allocations in the regulated Murrumbidgee Valley, NSW. DPI Water (2015), NSW, How water is shared in the regulated Murrumbidgee Valley, NSW. DPI Water, 2015, Guide to the components of a water access licence certificate, NSW.

Dutra, E., Balsamo, G., Calvet, J.-C., Minvielle, M., Eisner, S., Fink, G., Pessenteiner, S., Orth, R., Burke, S., van Dijk, A. I. J. M., Polcher, J., Beck, H. E. and Martinez de la Torre, A.: Report on the current state-of-the-art Water Resources Reanalysis. [online] Available from: http://earth2observe.eu/files/Public Deliverables/D5.1_Report on the WRR1 tier1.pdf, 2015.

GIWP. 2007. Economic Value of Water and Policy Intervention in the Hai Basin. Beijing, Ministry of Water Resources.

Green D., Petrovic J., Moss P., Burrell M. (2011) Water resources and management overview: Murrumbidgee catchment, NSW Offce of Water, Sydney (https://www.usbr.gov/newsroom/newsrelease/detail.cfm?RecordID=61677 accessed on 14 March 2018.) Irrigation corporations – information on water access licences, 2013 (http://www.water.nsw.gov.au/water-licensing/corporate- licences/irrigation-corporations/information-on-water-access-licences)

J. Baker J. Van Sickle D. White ;2002, Water Sources and Allocation, Willamette River Basin planning Atlas.

Kaune, Alexander & Werner, Micha & Rodríguez, Erasmo & Karimi, Poolad & Fraiture, Charlotte. (2017). A novel tool to assess available hydrological information and the occurrence of sub-optimal water allocation decisions in large irrigation districts. Agricultural Water Management.

Khan, S; Rana, T; Beddek, R; Paydar, Z; Carroll, Jason; Blackwell, J. Whole of catchment water and salt balance to identify potential water saving options in the Murrumbidgee Catchment. Griffith: CSIRO; 2004.

Krysanova, V., Vetter, T., Eisner, S., Huang, S., Pechlivanidis, I., Strauch, M., ... & Chamorro, A. (2017). Intercomparison of regional-scale hydrological models and climate change impacts projected for 12 large river basins worldwide—a synthesis. Environmental Research Letters, 12(10), 105002.

Le Quesne, T., Pegram, G., & von der Heyden, D. (2007). Allocating Scarce Water: A primer on Water allocation, water rights and water markets. WWF Water security Series 1. Godalming. UK.WWF-UK. http://assets.wwf.org.uk/downloads/scarce_water.pdf

Lerma - Chapala Basin Case Study: a fruitful sustainable water management experience, National Water Commission of Mexico (CONAGUA), September, 2011

López, P. L., Sutanudjaja, E. H., Schellekens, J., Sterk, G., & Bierkens, M. F. (2017). Calibration of a large-scale hydrological model using satellite-based soil moisture and evapotranspiration products. Hydrology and Earth System Sciences, 21(unknown), 3125-3144.

Mark Moore, 2009, NSWIC, ENTITLEMENT CONVERSION POLICY, 100311

Masafu, C. K., Trigg, M. A., Carter, R., & Howden, N. J. (2016). Water availability and agricultural demand: An assessment framework using global datasets in a data scarce catchment, Rokel-Seli River, Sierra Leone. Journal of Hydrology: Regional Studies, 8, 222-234

MCIP, Murrumbidgee catchment irrigation profile, 2003, NSW Agriculture,

Milly, P. C. D., K. A. Dunne, and A. V. Vecchia, 2005: Global pattern of trends in streamflow and water availability in a changing climate. Nature, 438, 347–350, doi:10.1038/nature04312.

Molden, D., 2007. Pathways for increasing agricultural water productivity.

Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, R. D., & Veith, T. L. (2007). Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE, 50(3), 885-900.

Morris, B L, Lawrence, A R L, Chilton, P J C, Adams, B, Calow R C and Klinck, B A. (2003) Groundwater and its Susceptibility to Degradation: A Global Assessment of the Problem and Options for Management. Early Warning and Assessment Report Series, RS. 03-3. United Nations Environment Programme, Nairobi, Kenya

Murrumbidgee Catchment Action Plan CAP, 2013

Murrumbidgee Irrigaton Limited (MI). 2017. company overview, www.mirrigaton.com.auwww.mirrigaton.com.au

Murrumbidgee River Basin (NSW Real Time Data) (http://realtimedata.water.nsw.gov.au) Murrumbidgee Water Resource Plan (Surface Water)- February 2017, (www.dpi.nsw.gov.au)

Murrumbidgee Water Resource Plan (Surface Water, SW9)- February 2017, (http://www.water.nsw.gov.au/ data/assets/pdf_file/0007/698155/Status-and- Issues-Paper-Murrumbidgee-SW-WRP_170217_final_2.pdf)

Patrick, M. J. (2012). Scale and justice in water allocation. Retrieved from http://ro.ecu.edu.au/theses/474

Petra Hellegers & Xavier Leflaive (2015) Water allocation reform: what makes it so difficult?, Water International, 40:2, 273-285, DOI: 10.1080/02508060.2015.1008266

Qureshi, M. E., Ranjan, R. and Qureshi, S. E. (2010), An empirical assessment of the value of irrigation water: the case study of Murrumbidgee catchment*. Australian Journal of Agricultural and Resource Economics, 54: 99–118. doi:10.1111/j.1467- 8489.2009.00476.x

R. Speed, Li Y., T. Le Quesne, G. Pegram and Z. Zhiwei (2013) Basin Water Allocation Planning. Principles, procedures and approaches for basin allocation planning, UNESCO, Paris

R.T. Kingsford and R.F. Thomas (2001), Changing water regimes and wetland habitat on the Lower Murrumbidgee floodplain of the Murrumbidgee River in arid Australia.

Ramsar handbooks for the wise use of wetlands, 2010

Report on the verification and value that the datasets developed in EartH2Observe have in supporting operational and local decisions, 2017, EartH2Observe

Resource Assessment & Allocations, Murrumbidgee Valley, July 2016, (NSW, 2016a)

Ribbons, C., 2009, Water availability in New South Wales Murray-Darling Basin regulated rivers, NSW Department of Water and Energy, Sydney

Rice growing guide 2016–17, NSW Department of Industry, Skills and Regional Development.

S Khan, S Mushtaq, A Ahmad & M Hafeez (2008) Trade-off analysis for restoring environmental flows through irrigation demand management, Australian Journal of Water Resources, 12:1, 1-20

Schellekens, J., Dutra, E., Martínez-De La Torre, A., Balsamo, G., Van Dijk, A., Sperna Weiland, F., Minvielle, M., Calvet, J. C., Decharme, B., Eisner, S., Fink, G., Flörke, M., Peßenteiner, S., Van Beek, R., Polcher, J., Beck, H., Orth, R., Calton, B., Burke, S., Dorigo, W. and Weedon, G. P.: A global water resources ensemble of hydrological models: The eartH2Observe Tier-1 dataset, Earth Syst. Sci. Data, 9(2), 389–413, doi:10.5194/essd-9-389-2017, 2017

Snowy Water Inquiry, 1998. Snowy Water Inquiry Issues Paper. Snowy Water Inquiry, Governor Macquarie Tower, Sydney.

Stanski, H. R., Burrows, W. R., & Wilson, L. J. (1989). Survey of common verification methods in meteorology.

Sustainable Management of Water Resources in Agriculture, OECD, 2010

SW (State Water Corporation), Dam Facts and Figures, Blowering Dam, State Water Corporation, NSW, 2009b

SW (State Water Corporation),Dam Facts and Figures, Burrinjuck Dam, State Water Corporation, NSW, 2009a

Swainson, B.; de Loë, R.C. and Kreutzwiser, R.D. 2011. Sharing water with nature: Insights on environmental water allocation from a case study of the Murrumbidgee catchment, Australia.

Tom Le Quesne, Guy Pegram and Constantin Von Der Heyden Turral, H. N., T. Etchells, H. M. M. Malano, H. A. Wijedasa, P. Taylor, T. A. M. McMahon, and N. Austin (2005), Water trading at the margin: The evolution of water markets in the Murray-Darling Basin, Water Resour. Res., 41, W07011, doi:10.1029/2004WR003463.

United Nations (2000) Principles and practices of water allocation among water-user sectors. Water Resources Series No. 80. United Nations, New York

van Dijk, A. I. J. M., Renzullo, L. J., Wada, Y., & Tregoning, P. (2014). A global water cycle reanalysis (2003– 2012) merging satellite gravimetry and altimetry observations with a hydrological multi-model ensemble. Hydrol. Earth Syst. Sci.,, 18, 2955–2973. doi: doi:10.5194/hess-18-5-2955-2014

Water availability in the Murrumbidgee. A report to the Australian Government from the CSIRO Murray-Darling Basin Sustainable Yields Project. CSIRO, Australia. 155pp.

Water sharing in the Murrumbidgee Regulated River Progress report 2004 to 2008, May 2009, NSW Department of Water and Energy (DWE, 2009)

Water Sharing Plan for the Murrumbidgee Regulated River Water Source 2016, Water Sharing Plan for the Murrumbidgee Regulated River Water Source 2003

WMO. (2012). Climate and Meteorological Information Requirements for Water Management: A Review of Issues. URL: http://goo.gl/cQTqoL (accessed 05.01.15.)

Appendices

Appendices A Water Allocation Data (2014-15)

Abbildung in dieser Leseprobe nicht enthalten

Appendices B Example of Water Access License in NSW

Abbildung in dieser Leseprobe nicht enthalten

Appendices C Longitudinal Profile of the Murrumbidgee

Abbildung in dieser Leseprobe nicht enthalten

Appendices D Model Input data

Abbildung in dieser Leseprobe nicht enthalten

Final del extracto de 115 páginas

Detalles

Título
Assessment of consistency in water allocation decision making
Subtítulo
The case of the Murrumbidgee Regulated River in Australia
Universidad
Delft University of Technology  (UNESCO-IHE, the Netherlands)
Curso
Hydrology and Water Resources
Calificación
9.00/10.00
Autor
Año
2018
Páginas
115
No. de catálogo
V459972
ISBN (Ebook)
9783668906563
ISBN (Libro)
9783668906570
Idioma
Inglés
Palabras clave
Water allocation, Murrumbidgee, General security, Murray Darling, consistency, decision making, Burrinjuck, Blowering, rice production, availability of water, New South Wales, NSW, Australia
Citar trabajo
Mohammad Faysal Chowdhury (Autor), 2018, Assessment of consistency in water allocation decision making, Múnich, GRIN Verlag, https://www.grin.com/document/459972

Comentarios

  • No hay comentarios todavía.
Leer eBook
Título: Assessment of consistency in water allocation decision making



Cargar textos

Sus trabajos académicos / tesis:

- Publicación como eBook y libro impreso
- Honorarios altos para las ventas
- Totalmente gratuito y con ISBN
- Le llevará solo 5 minutos
- Cada trabajo encuentra lectores

Así es como funciona