Hydrologic Modelling for the Northern Basin Review

January 2016

Published by the Murray–Darling Basin Authority GPO Box 1801, Canberra ACT 2601 MDBA publication no.: 35/16 02 6279 0100 ISBN (online): 978-1-921914-71-3 [email protected] © Murray–Darling Basin Authority 2016 With the exception of the Commonwealth Coat of Arms, mdba.gov.au the MDBA logo, trademarks and any exempt photographs and graphics (these are identified), this publication is provided under a Creative Commons Attribution 4.0 licence.

https://creativecommons.org/licenses/by/4.0 The Murray‒Darling Basin Authority’s preference is that you attribute this publication (and any Murray‒ Darling Basin Authority material sourced from it) using the following wording within your work: Title: Hydrologic modelling for the Northern Basin Review Source: Licensed from the Murray‒Darling Basin Authority under a Creative Commons Attribution 4.0 Licence Accessibility The Murray‒Darling Basin Authority makes its documents and information available in accessible formats. On some occasions the highly technical nature of the document means that we cannot make some sections fully accessible. If you encounter accessibility problems or the document is in a format that you cannot access, please contact us.

Acknowledgement of the Traditional Owners of the Murray–Darling Basin The Murray–Darling Basin Authority acknowledges and pays respect to the Traditional Owners, and their Nations, of the Murray–Darling Basin, who have a deep cultural, social, environmental, spiritual and economic connection to their lands and waters. The MDBA understands the need for recognition of Traditional Owner knowledge and cultural values in natural resource management associated with the Basin. The approach of Traditional Owners to caring for the natural landscape, including water, can be expressed in the words of the Northern Basin Aboriginal Nations Board: As the First Nations peoples (Traditional Owners) we are the knowledge holders, connected to Country and with the cultural authority to share our knowledge. We offer perspectives to balance and challenge other voices and viewpoints. We aspire to owning and managing water to protect our totemic obligations, to carry out our way of life, and to teach our younger generations to maintain our connections and heritage through our law and customs. When Country is happy, our spirits are happy. This report may contain quotes by Aboriginal people who have passed away. The use of terms ‘Aboriginal’ and ‘Indigenous’ reflects usage in different communities within the Murray–Darling Basin. Published by the Murray‒Darling Basin Authority Postal Address: GPO Box 1801, Canberra ACT 2601 Telephone: (02) 6279 0100 international + 61 2 6279 0100 Facsimile: (02) 6248 8053 international + 61 2 6248 8053 Email: [email protected] Internet: www.mdba.gov.au Cover image: Darling River near Bourke, NSW (photo by Josh Smith Photography)

Hydrologic Modelling for the Northern Basin Review

Contents Executive Summary ...... 1 1. Introduction ...... 4 1.1. Northern Basin Review ...... 4 1.2. Basin Plan Modelling ...... 5 1.3. The Need for Additional Northern Basin Modelling ...... 9 1.4. Model Assumptions and Uncertainties ...... 10 1.4.1. Inherent Model Uncertainty ...... 10 1.4.2. Basin Plan Model Uncertainty ...... 11 1.5. Scope of this Report ...... 11 2. Hydrologic Modelling Framework ...... 13 2.1. Queensland Models ...... 16 2.1.1. Paroo Model ...... 16 2.1.2. Moonie Model...... 16 2.1.3. Warrego Model ...... 17 2.1.4. Nebine Model ...... 17 2.1.5. Condamine-Balonne Model ...... 17 2.1.6. Border Rivers ...... 18 2.2. New South Wales Models ...... 19 2.2.1. Namoi ...... 19 2.2.2. Gwydir ...... 19 2.2.3. Macquarie-Castlereagh ...... 20 2.2.4. Barwon-Darling ...... 21 2.3. Without Development Scenario ...... 22 2.4. Baseline Scenario ...... 22 2.5. Basin Plan Scenarios ...... 22 3. Flow and Connectivity Through the Northern Basin ...... 24 3.1. Climate, Rainfall, Flow and River Regulation ...... 24 3.2. Connectivity ...... 26 3.3. Quantifying Northern Basin Connectivity ...... 26 4. Modelling Approach ...... 32 4.1. Condamine–Balonne Modelling ...... 32 4.2. Whole-of-north Modelling ...... 38 5. Modelling Methodology ...... 44 5.1. Representing Water Recovery ...... 44

Page i

Hydrologic Modelling for the Northern Basin Review

5.1.1. Apportionment of the Shared Component ...... 44 5.1.2. Representing Water Recovery in the Models ...... 46 5.2. Representing Environmental Water Use ...... 49 5.2.1. Environmental Water Accounting in the Models ...... 50 5.2.2. Demand Series for Local Outcomes ...... 51 5.2.3. Demand Series for Downstream Outcomes ...... 59 5.2.4. Iterating a Model Scenario to Completion ...... 67 5.2.5. Protection of Water in the Barwon–Darling ...... 68 5.3. Model Output Analysis ...... 69 6. Condamine–Balonne Scenarios ...... 71 6.1. Model Package 1 — Location of Recovery ...... 74 6.1.1. Recovery Upstream of Beardmore Dam ...... 74 6.1.2. Flow to the Narran Lakes ...... 79 6.2. Model Package 2 — Entitlement Type Dependence ...... 81 6.2.1. Unregulated Entitlement Mix ...... 82 6.2.2. Regulated Recovery ...... 87 6.3. Model Package 3 — Horizontal Slicing ...... 89 6.4. Model Package 4 — SDL Sensitivity ...... 93 7. Whole-of-north Scenarios ...... 102 7.1. SDL Sensitivity ...... 102 7.2. Water Recovery and Use Sensitivity ...... 107 7.2.1. Geographic Location of Recovery ...... 107 7.2.2. Options for Delivering Water to the Barwon–Darling...... 115 7.3. Refined Scenarios ...... 119 7.3.1. Refined 321 GL Scenario ...... 120 7.3.2. Refined 345 GL Scenario ...... 125 8. Alternate Mechanisms Analysis...... 132 8.1. Methodology ...... 133 8.1.1. Threshold & Duration SFIs ...... 134 8.1.2. Volumetric SFIs ...... 136 8.1.3. Dry Spell SFIs ...... 136 8.2. Results and Discussion ...... 139 8.2.1. Barwon-Darling ...... 139 8.2.2. Condamine-Balonne...... 143 8.2.3. Namoi ...... 148

Page ii

Hydrologic Modelling for the Northern Basin Review

8.2.4. Border Rivers ...... 150 8.3. Alternate Mechanisms Conclusions ...... 151 9. Downstream Deliverability Analysis ...... 153 9.1. Demand Series Analysis ...... 154 9.2. Downstream Water Delivery ...... 155 9.2.1. Demand window analysis ...... 155 9.2.2. Event success analysis ...... 159 9.3. Downstream Deliverability Conclusions ...... 165 10. References ...... 167 Appendix A: SFI Results ...... 169 Appendix B: Baseflow Metric Results ...... 177 Appendix C: Mass Balance Tables and Graphs ...... 183

Page iii

Hydrologic Modelling for the Northern Basin Review

List of Figures Figure 1: Overview of the Northern Basin Review work program — the hydrologic modelling component is marked as the dashed outline section ...... 7 Figure 2: Example 12-month period of modelled daily flows. The red line traces the flow under baseline conditions (i.e. water sharing arrangements prior to the Basin Plan), while the blue region shows the flow added under one of the modelled Basin Plan options (in this case, 390 GL of water recovery). For reference, the dashed line shows the same period under without development conditions...... 8 Figure 3: Schematic diagram of the main rivers in the Northern Basin ...... 24 Figure 4: Ratio of active storage capacity to average annual inflow for the major public storages throughout the Northern Basin ...... 25 Figure 5: End of system flows for each catchment in the Northern Basin for both without development and baseline (pre-Basin Plan conditions) models ...... 27 Figure 6: Proportion of inflows reaching the end of system for each catchment in the Northern Basin for both without development and baseline (pre-Basin Plan conditions) models ...... 28 Figure 7: Examples of a Barwon–Darling event sourced primarily from Queensland catchments (upper panel) and NSW catchments (lower panel). The dotted line traces the flow at Bourke and the coloured lines trace the end of system flows from each catchment...... 29 Figure 8: Schematic representation of the average tributary contribution to a flow event at Bourke in the range 15,000 – 20,000 ML/d...... 30 Figure 9: Schematic diagram for the Condamine–Balonne; the Lower Balonne Floodplain and Narran Lakes areas are outlined in green, and the locations to monitor modelled environmental flows are marked in light blue ...... 33 Figure 10: Schematic map of the Condamine–Balonne including the six sub-regions considered for the spatial modelling work (from north to south, the six sub-regions are ‘Streams of the main river stem’, ‘Upstream Beardmore’, ‘St George’, ‘Jack Taylor Weir to Bifurcation 1’, ‘Narran System’, ‘Lower Balonne System’) ...... 34 Figure 11: Condamine–Balonne model scenarios completed as part of the Northern Basin Review. The aggregated bar shows the total water recovered in the catchment, sub-divided by colour to represent sub-region and entitlement type recovered...... 37 Figure 12: Schematic map of the Barwon–Darling system and its main tributaries ...... 40 Figure 13: Schematic map of the whole-of-north model scenarios completed for the Northern Basin Review. The vertical axis traces the Northern Basin recovery volume, while the horizontal axis traces the core modelling assumptions included in each scenario. A definition of the terms (e.g. ‘targeted’, ‘semi-targeted’) is provided in the text...... 41 Figure 14: Water recovery distributions represented in each of the NBR whole-of-north model scenarios (note that scenarios B and H both represented the same 390 GL recovery pattern) . 43 Figure 15: Example environmental demand series, showing a requested flow event in the Lower Namoi (measured at Bugilbone) over a 12-month period. The blue region shows the additional environmental flow requested for delivery to supplement existing flow events...... 54 Figure 16: Example demand series event for the Gwydir Wetlands, measured at Yarraman Bridge. The orange line traces the requested flows to supplement the initial trigger event...... 57 Figure 17: Decision tree determining events to be selected under Strategy 1 (coordinated releases) ...... 63

Page iv

Hydrologic Modelling for the Northern Basin Review

Figure 18 Stylised end of system hydrograph of uncoordinated downstream demand series generation in the Border Rivers. The shaded area represents the additional flow requested from storage...... 65 Figure 19 Schematic diagram for the Macquarie–Castlereagh ...... 66 Figure 20 Stylised hydrograph of catchment-scale (uncoordinated) demand series generation in the Macquarie Castlereagh Catchment. The shaded area represents the additional flow requested from storage...... 67 Figure 21: Schematic diagram for the Condamine–Balonne. The six sub-regions adopted for NBR modelling purposes are outlined in different colours...... 73 Figure 22: Differential flow duration curve at Brenda for the U/S Beardmore scenarios. The anchor points in the chart are baseline flows (0%) and flows under the 0 GL U/S Beardmore scenario (100%)...... 78 Figure 23: Differential flow duration curve at Wilby Wilby for the U/S Beardmore scenarios. The anchor points in the chart are baseline flows (0%) and flows under the 0 GL U/S Beardmore scenario (100%)...... 79 Figure 24: Relationship between water recovery upstream of Bifurcation 1 and increase in flows to Narran Lakes...... 81 Figure 25: Split between water harvester and overland flow recovery for the three entitlement scenarios...... 84 Figure 26: Differential flow duration curve at Brenda (Culgoa River) for the three entitlement scenarios. This diagram is anchored by baseline flow at 0%...... 85 Figure 27: Differential flow duration curve at Wilby Wilby (Narran River) for the three entitlement scenarios. This diagram is anchored by baseline flow at 0%...... 86 Figure 28 Example supplemented low flow demand response ...... 88 Figure 29: Differential flow duration curve at Brenda for the horizontal slicing scenarios. The anchor points in the chart are baseline flows (0%) and flows under the Qld 100 GL option (100%; scenario 1041)...... 92 Figure 30: Differential flow duration curve at Wilby Wilby for the horizontal slicing scenarios. The anchor points in the chart are baseline flows (0%) and flows under the Qld 100 GL option (100%; scenario 1041)...... 93 Figure 31: SDL options modelled for the Condamine–Balonne. The aggregated bar shows the total water recovered in the catchment, sub-divided by colour to represent sub-region and entitlement type recovered...... 95 Figure 32: Differential flow duration curve at Brenda for the SDL sensitivity scenarios. The anchor point in this chart is baseline flows at 0%...... 97 Figure 33: Differential flow duration curve at Wilby Wilby for the SDL sensitivity scenarios. The anchor point in this chart is baseline flows at 0%...... 98 Figure 34: Differential flow duration curve for end of system flows under the Condamine– Balonne SDL sensitivity scenarios. The anchor point in this chart is baseline flows at 0%...... 99 Figure 35: Relationship between water recovery volume and average increase in end-of-system flows for the Condamine–Balonne ...... 101 Figure 36: Differential flow duration curve at Bourke for the five SDL sensitivity scenarios. The anchor point in the chart is baseline flow (0%)...... 105 Figure 37: Comparison between baseline and Scenario B (390 GL) of peaks of flow events at Bourke. The events have been divided into 1,000 ML/d increments. The dashed line traces a 1- to-1 relationship (i.e. no change from baseline flow)...... 106

Page v

Hydrologic Modelling for the Northern Basin Review

Figure 38: Differential flow duration curve at Bourke for the 350 GL and 390 GL scenarios. The anchor point in the chart is baseline flows (0%)...... 109 Figure 39: Differential flow duration curve for inflows to the Barwon–Darling for the two 320 GL scenarios. The anchor point in the chart is baseline flow (0%)...... 113 Figure 40: Differential flow duration curve at Bourke for the two 320 GL scenarios. The anchor point in the chart is baseline flow (0%)...... 114 Figure 41: Differential flow duration curve at Bourke for the 278, 320 and both 390 GL scenarios. The anchor point in the chart is baseline flows (0%)...... 118 Figure 41 Differential flow duration curves showing the change in inflows to the Barwon–Darling from (a) the Condamine–Balonne Bourke, (b) the Macquarie–Castlereagh, and (c) all tributaries combined. The anchor point in these charts is baseline flow (0%)...... 123 Figure 42: Differential flow duration curve at Bourke for the 320 GL and 321 GL (refined) scenarios. The anchor point in the chart is baseline flow (0%)...... 124 Figure 43 Differential flow duration curves showing the change in inflows to the Barwon–Darling from (a) the Condamine–Balonne Bourke, (b) Border Rivers, (c) Namoi, and (d) Macquarie– Castlereagh. The anchor point in these charts is baseline flow (0%)...... 128 Figure 44: Differential flow duration curve for total Barwon–Darling inflows. The anchor point in the chart is baseline flow (0%)...... 129 Figure 45: Differential flow duration curve for total flows at Bourke. The anchor point in the chart is baseline flow (0%)...... 130 Figure 46 Stylised diagram of the Northern Basin, where the sites used for testing event enhancement opportunities are marked in light blue...... 133 Figure 47 Example SFI shortfall hydrograph ...... 135 Figure 48 Example Narran Lakes event ...... 136 Figure 49 Stylised diagram of the Condamine-Balonne Catchment ...... 137 Figure 50 Example unregulated cease to flow event in the Culgoa River ...... 138 Figure 51 Example regulated fresh in the Culgoa River...... 138 Figure 52: Bourke SFI shortfall analysis results ...... 139 Figure 53 Louth SFI shortfall analysis results ...... 140 Figure 54 Wilcannia SFI shortfall analysis results ...... 141 Figure 55: Brenda SFI shortfall analysis results...... 144 Figure 56 Brenda SFI shortfall analysis results with annual Lower Balonne diversions ...... 145 Figure 57 Culgoa freshes analysis results ...... 146 Figure 58 Narran Lakes SFI shortfall analysis results ...... 147 Figure 59 Bugilbone SFI shortfall analysis results ...... 148 Figure 60 Bugilbone SFI shortfall analysis results with annual Namoi diversions ...... 149 Figure 61: Mungindi SFI shortfall analysis results ...... 150 Figure 62 Mungindi SFI shortfall analysis results with annual Border Rivers diversions ...... 151 Figure 63: Stylised representation of northern Basin catchments ...... 154 Figure 64 Summary of downstream demands ...... 155 Figure 65 Stylised representation of the Namoi River Catchment ...... 156 Figure 66 Example demand response from the Namoi model ...... 156 Figure 67 Average additional flow in each catchment during demand windows ...... 157 Figure 68 Stlysied representation of the Macquaire-Castlereagh catchment ...... 158 Figure 69 Additional inflows to the Barwon-Darling during demand windows, from Baseline to Northern Standard ...... 159 Figure 70 Example successfully delivered event at Bourke ...... 160

Page vi

Hydrologic Modelling for the Northern Basin Review

Figure 71 Example failed event at Bourke, the targeted SFI is 6,000 ML/d at Bourke for 14 days ...... 161 Figure 72 Average chane in flow during downstream demand windows...... 163 Figure 73 Demand activation porportions ...... 164

Page vii

Hydrologic Modelling for the Northern Basin Review

List of Tables Table 1: Comparison of SFI achievement for the baseline scenario (845) and the updated models provided by NSW ...... 15 Table 2: Comparison of long-term average diversions for the existing baseline model and the updated models provided by NSW ...... 16 Table 3: Summary of Condamine–Balonne models ...... 18 Table 4: Breakdown of baseline diversions and water recovery by sub-catchment for all Basin Plan Condamine–Balonne model scenarios; the grey row indicates the sub-region that was not considered for recovery as part of this study. Model scenarios are divided into parameter of interest (e.g. ‘Initial SDL Scenarios’) and by model run number (e.g. 980)...... 36 Table 5: Catchment water recovery distributions represented in the Northern Basin Review model scenarios; grey cells indicate catchment recoveries that are unchanged from current recovery volumes and were not altered through the modelling work program...... 42 Table 6: In-valley and existing recovery volumes for Northern catchments ...... 46 Table 7: Site-specific flow indicators underlying the demand series for local needs ...... 51 Table 8: Seven-day event volumes used to trigger environmental releases for each Gwydir wetlands SFI ...... 56 Table 9: Number of events included in the 114-year demand series for local environmental requirements in each scenario ...... 58 Table 10: End of system locations and maximum target flows for the regulated catchments in both the coordinated and not coordinated model scenarios ...... 60 Table 11: SFIs in the Barwon–Darling — indicators targeted through the Strategy 1 (coordinated releases) are coloured green...... 61 Table 12: Low flow targets listed in the Barwon–Darling water sharing plan under the Interim Unregulated Flow Management Plan for the North West ...... 64 Table 13: Recovery pattern for the Upstream Beardmore scenarios ...... 75 Table 14: Long-term average inflow to Beardmore, and flow at Brenda and Wilby Wilby under baseline conditions and for the ‘U/S Beardmore’ scenarios ...... 77 Table 15: Model scenarios used to examine the relationship between water recovery and flow to Narran Lakes ...... 80 Table 16: Model parameters for the NBR entitlement type scenarios ...... 83 Table 17: Long-term average flows at Brenda (Culgoa River) and Wilby Wilby (Narran River) for the three entitlement scenarios ...... 85 Table 18 Condamine-Balonne regulated recovery response ...... 88 Table 19: Recovery pattern for the horizontal slicing scenarios ...... 90 Table 20: Long-term average flows at Brenda (Culgoa River) and Wilby Wilby (Narran River) for the horizontal slicing scenarios ...... 91 Table 21: Long-term average flows through the Lower Balonne for the SDL sensitivity scenarios ...... 96 Table 22: Long-term average flow results for the five model scenarios testing SDL sensitivity 104 Table 23: Average increased in peak flow for each event in the four SDL sensitivity scenarios ...... 107 Table 24: Long-term average flow results for the model scenarios testing recovery location across the Northern Basin ...... 108 Table 25: Differences in water recovery distribution for the two 320 GL options (note that all values have been rounded to integers). Grey rows indicate regions that were excluded from

Page viii

Hydrologic Modelling for the Northern Basin Review further contribution towards the shared recovery component. Green cells indicate catchments with an increase in recovery from Scenario E, orange indicates a decrease...... 110 Table 26: Differences in assumed water use ...... 116 Table 27: Long-term average flow results for the model scenarios testing water delivery options across the Northern Basin ...... 117 Table 26: Catchment water recovery volumes for Scenario J (321 GL refined), and a summary of the contributing information from previous model scenarios ...... 121 Table 27: Long-term average flow results for the refined model scenarios ...... 122 Table 28: Catchment water recovery volumes for Scenario I (345 GL refined), and a summary of the contributing information from previous model scenarios ...... 126 Table 29: Comparison of catchment-scale recovery volumes for the 321 and 345 GL refined scenarios...... 127 Table 30 SFIs analysed for event enhancement opportunities ...... 134 Table 31 Barwon-Darling SFI Shortfall Analysis Results ...... 142 Table 32 Barwon-Darling SFI Shortfall Analysis Results in terms of Annual Diversions ...... 143 Table 33 Brenda SFI shortfall analysis results ...... 144 Table 34 Brenda results as a proportion of annual diversions ...... 145 Table 35 Culgoa regulated freshes analysis results ...... 146 Table 36 Narran Lakes SFI shortfall analysis results ...... 147 Table 37 Bugilbone SFI shortfall analysis results ...... 148 Table 38 Bugilbone results as a proportion of annual Namoi diversions ...... 149 Table 39: Mungindi SFI shortfall analysis results ...... 150 Table 40 Border Rivers results as a proportion of annual diversions ...... 151 Table 41 Average additional flow in each catchment during demand windows ...... 157 Table 42 Requested regulated events success rates ...... 162

Page ix

Hydrologic Modelling for the Northern Basin Review

Executive Summary Northern Basin Review The Basin Plan sets limits on the amount of water that can be extracted for consumptive use in order to achieve a healthy, working Murray–Darling Basin. These sustainable diversion limits (SDLs) are set to ensure water resources in the basin continue to support strong and vibrant communities, resilient industries, including food and fibre production, and a healthy environment. On 22 November 2016 the MDBA released the findings from the Northern Basin Review. The purpose of the review was re-examine the SDLs set with the Basin Plan in 2012. The review was adopted with the recognition that the 2012 decision was based on knowledge that, while the best available at the time, could be improved through further research and ongoing community consultation. The findings from the review recommended a reduction in the total volume of water to be recovered for the environment across the Northern Basin. In light of the updated research and local community feedback, the MDBA are recommending that the volume set in 2012 (390 GL) be reduced to 320 GL provided there are commitments from governments to implement a range of complementary ‘toolkit measures’. The Role of Hydrologic Modelling This report describes the extensive three-year hydrologic modelling work program undertaken for the Northern Basin Review. The purpose of this work was to measure the long-term flow changes that could occur under a wide range of possible Basin Plan options under consideration by the Authority. For each option, the changes in flow were translated into a set of outcomes using the social, economic and environmental research streams, also part of the Northern Basin Review. The various options were compared through a triple-bottom line assessment framework which incorporated new research and knowledge, while stakeholder and jurisdictional consultation also provided valuable input. The final proposed water recovery volumes for the northern basin therefore represent a considered, evidence-based, triple bottom line judgement call. The proposed water recovery volumes and rationale for these are available in the Northern Basin Review report (MDBA 2016a). The primary task of the hydrologic modelling work program was to test a number of SDL options. For this purpose, a relatively large number of model scenarios were completed representing Northern Basin recovery volumes ranging from 278 GL (estimated current recovery) to 415 GL. An examination of the outcomes associated with each scenario helped the Authority to understand how industries, communities and the environment respond to changes in flow. Toolkit Measures The Authority recognises that mechanisms other than water recovery will be required to achieve the desired outcomes of the Basin Plan. The decision to reduce the recovery volume from 390 GL to 320 GL therefore includes a provision that the Australian, Queensland and New South Wales governments commit to the implementation of the toolkit measures. This report includes an analysis of those toolkit measures which are related to flow (i.e. hydrological measures):

Page 1

Hydrologic Modelling for the Northern Basin Review

 The modelling scenarios revealed the need for environmental flow protection through the Condamine–Balonne if recovery upstream of Beardmore Dam was pursued.  Analysis of the modelling data demonstrated that the location of water recovery (not just the volume) is an important factor influencing the flows achieved by the Basin Plan, especially through the Condamine–Balonne and Barwon–Darling river systems.  Event-based mechanisms (such as temporary trade) can benefit some environmental assets, including the Narran Lakes and Lower Balonne Floodplain.  The coordination of environmental releases from storages across the Northern Basin can be used to augment environmental outcomes in the Barwon–Darling. Modelling Platform The models that underpin the Basin Plan modelling framework were provided by the Basin state governments, and have been the subject of development, calibration and testing for many years. The MDBA and Basin state governments acknowledge that the models contain inherent uncertainties, which are generally determined by the extent of available calibration data (i.e. gauged flows, diversions, storage releases and spills, evaporation, rainfall, and so on). It is also acknowledged that there is generally less calibration data at hand across the northern Basin compared with catchments further south, and that the northern catchments also have a more variable climate, both of which affect the model calibration. Despite these acknowledged uncertainties, independent reviews have confirmed that the modelling platform used for the Northern Basin Review is fit-for-purpose1. Hydrologic modelling of the Basin Plan necessarily includes an element of forecasting. The Basin Plan has not been fully implemented, and many of its important elements are still undergoing development. Notably, the pattern of any future water recovery is not certain, and the strategies underlying future environmental water use across the Northern Basin are still being developed. To include these elements in the model, the MDBA was required to anticipate the outcomes of ongoing processes. The models therefore included estimates of the future patterns of water recovery and use. Where possible, these estimates were informed by actions that have been in practice. Also, a number of exploratory scenarios were completed to measure the materiality of these assumptions. For this reason, the models do not predict the outcomes of the Basin Plan. They are instead an indication of the outcomes that could be achieved. The actual outcomes will depend on future policy choices to be made as the Basin Plan continues to be implemented. Recommended Option The current water recovery target in the Northern Basin is 390 GL, consisting of local reductions from each Northern catchment (totalling 247 GL), plus an additional 143 GL shared recovery to achieve environmental outcomes in the Barwon–Darling.

1 Podger, GM, Barma, D, Neal, B, Austin, K and Murrihy, E (2010), River System Modelling for the Basin Plan Assessment of fitness for purpose. CSIRO: Water for a Healthy Country National Research Flagship, Canberra, December 2010 Bewsher (2016) Review of the Hydrological Modelling Frameworks used to inform Potential Basin Plan Amendments, prepared for the Murray–Darling Basin Authority, 2016

Page 2

Hydrologic Modelling for the Northern Basin Review

The Authority have recommended a reduced recovery volume of 320 GL, consisting of an overall increase in local reductions (279 GL) plus an additional 41 GL shared recovery. The increase in local reductions reflects the updated knowledge on how local environmental needs are better met with in-catchment recovery. The catchment-scale breakdown of recovery volumes is presented in Table E.1.

Table E.1: The water reductions proposed by the Authority

Proposed Local Assumed Shared Total Reduction Catchment Reduction (GL) Reduction (GL) (GL)

Paroo 0 0 0

Warrego 8 0 8 Nebine 1 0 1 Moonie 0 2 2 Condamine-Balonne 100 0 100 QUEENSLAND Queensland Border Rivers 14 15 29 Queensland Total 123 17 140

Intersecting Streams 0 8 8 Gwydir 42 0 42 NSW Border Rivers 7 0 7 Namoi 20 0 20 Macquarie-Castlereagh 55 16 71 Barwon-Darling 32 0 32

NEW WALES SOUTH NSW Total 156 24 180 Whole of North Total 279 41 320

The Authority have proposed a specific distribution of the 41 GL shared reduction, however both the NSW and Queensland governments can advise on an alternative distribution within their states. The proposed split listed in Table E.1 is based on the new knowledge gathered as part of the review. Of particular importance was the modelling analysis (described in this report) which quantified the relative ability of each catchment to contribute downstream flows into the Barwon–Darling. The recommendation listed in Table E.1 is expected to enhance Barwon– Darling flows with a relatively high level of efficiency. This report describes the model scenarios that were provided as an input to the Authority triple- bottom line decision making process. The 320 GL option recommended by the Authority is not provided as a model scenario in this report, but most of its aspects were drawn from existing scenarios. Hence the outcomes in the tributary catchments can largely be drawn from the scenarios described in this report. Outcomes in the Barwon–Darling are expected to be similar to those achieved in the other 320 GL scenarios described in this report, noting that flows may be further enhanced due to the choice of a relatively efficient shared reduction distribution.

Page 3

Hydrologic Modelling for the Northern Basin Review

1. Introduction In 2012, the Federal Government adopted the Basin Plan, providing the first integrated framework for water planning in the Murray–Darling Basin. The Basin Plan aims to restore healthy river systems for the benefit of the environment, communities and agriculture. A central component of the Basin Plan is the establishment of sustainable diversion limits (SDLs) that specify the maximum amount of water that can be taken from surface and ground water resources. At the Basin-scale, SDLs restrict the consumption of surface water resources to a long term average volume of 10,873 GL per year, requiring the recovery of 2,750 GL of water for the environment. Of this recovery volume, 390 GL are to be sourced from the Northern Basin. The SDLs will take effect in 2019 when Basin Plan compliant water resource plans are passed into legislation in each Basin state. The 390 GL recovery volume includes local reductions from each Northern catchment (totalling 247 GL), plus an additional 143 GL shared recovery to achieve environmental outcomes in the Barwon–Darling. The Basin Plan includes a default method for distributing this shared component (based on the relative baseline diversions of each catchment), however it also provides Basin governments flexibility to choose an alternative distribution. The SDLs adopted in 2012 were informed by a number of information streams, including an assessment of environmental, social and economic outcomes. Underlying this information were the results of a MDBA hydrologic modelling work program that commenced in 2009 and was reviewed and refined through to 2012. The purpose of the hydrological modelling work program was to inform the decision on the SDLs by demonstrating the ecological outcomes that could be achieved with the water recovery program under the Basin Plan. 1.1. Northern Basin Review Due to differences in geography, rainfall, and water management, surface water resources in the Northern Basin have been developed and managed differently to the Southern Basin — the proportion of Northern Basin flows regulated by dams is much lower, and many of the water licences in these regions are associated with unregulated conditions. As a result, the implementation of the Basin Plan in the Northern Basin will require a different approach. Upon finalising the Basin Plan, the MDBA agreed to do further research and investigations on the SDLs of the Northern Basin through the review provisions in Chapter 6.06 of the Basin Plan. This review would then determine if new information indicates a case for changing the SDLs recommended in 2012. While the science was the best available at the time, the MDBA felt a review provided a chance to fill some of the identified information gaps. This review encompasses the Northern Basin region as a whole, with a focus on the Condamine–Balonne and Barwon–Darling catchments. The Northern Basin Review (MDBA 2016a) began in 2013 and was supported by all Basin Water Ministers. An overview of the work program adopted for the review is shown in Figure 1. The review had two broad streams of work:  A research program to incorporate new knowledge and provide updated information to the Authority for their re-examination of Basin Plan settings in the Northern Basin  A consultation program with Northern Basin communities and jurisdictions to incorporate local views regarding the ongoing implementation of the Basin Plan

Page 4

Hydrologic Modelling for the Northern Basin Review

The research program included three themes, with the aim of providing updated information to the Authority for their re-examination of Basin Plan settings in the Northern Basin. The three branches of work were:  Environmental science program — to help better understand the needs of birds, fish and floodplain vegetation in the Barwon–Darling and Condamine–Balonne river systems. This included new research and review of the most current knowledge on ecological responses to flows.  Social and economic assessment — to explore likely impacts of water recovery on communities and industries. This included economic modelling to define the relationships between irrigated area and employment at the farm level, in related farm supply industries and in the non-agricultural private sectors for communities across the northern Basin.  Updated water recovery modelling — to evaluate the hydrological outcomes associated with alternative SDL settings for the Basin Plan, and to inform the environmental, social and economic assessments. The overall purpose of the research program was to investigate the environmental, social and economic outcomes associated with a number of water recovery scenarios, as an alternative to the SDLs set by the Basin Plan. An outline of the NBR process is provided in Figure 1. The role of the modelling component was to simulate hydrological conditions of a range of Basin Plan water recovery options. The outputs from the hydrologic modelling were then further examined using the social, economic and environmental assessment frameworks developed for the Northern Basin Review. This provided a substantial set of outcomes related to each scenario, allowing the Authority to determine the relative merits of each Basin Plan option through a triple-bottom line assessment tool (MDBA 2016b). 1.2. Basin Plan Modelling During 2009/10, the MDBA and CSIRO collaboratively built a whole-of-Basin hydrologic modelling framework to assist with the development of the Basin Plan (see Podger et al. 2010a and 2010b for further details). This framework links the 24 individual catchment models developed by the States, the MDBA and Snowy Hydro Limited, allowing a complete examination of the surface water resources in the Murray–Darling Basin. Basin states provided two model scenarios as a starting point:  Without development scenario — all aspects of development of water resources have been removed from the model, including infrastructure and consumptive use. This is the best available estimate of the natural river system but without accounting for land use changes and on-farm development.  Baseline scenario — the best estimate of water management operations prior to the Basin Plan. This scenario includes all entitlements, water allocation policies, water sharing rules, operating rules and infrastructure such as dams, locks and weirs as of June 2009 or as specified in Schedule 3 of the Basin Plan. Jointly, these two scenarios reveal the impact of water resource development (to 2009) on the flow regime of the Murray–Darling Basin.

Page 5

Hydrologic Modelling for the Northern Basin Review

The general modelling approach adopted for Basin Plan development from 2009 to 2012 was to modify the baseline models to produce a set of specific Basin Plan scenarios. This approach was adopted for Basin Plan development (i.e. from 2009 to 2012) and was a key component of the method used to determine the environmentally sustainable level of take (ESLT; MDBA 2011). A description of the modelling contribution to the ESLT determination is given by MDBA (2012a). The same approach was adopted for the Northern Basin Review, but it was updated to include new environmental science, and a larger set of modelled scenarios were completed. For each option, the models were modified to represent a different volume of recovered water, or an alternative option for managing environmental entitlements. Under all scenarios the modelled river system received the same series of daily inflow, rainfall and evaporation data over the 114-year modelling period (1895 – 2009). The only difference between each scenario was the input settings defining the modelled behaviour of the system. Once completed, each scenario provided a large dataset of daily time series, which showed the day-to-day behaviour of the river system, where this ‘behaviour’ included aspects such as flows in the river on each day or the volume of water held in public storages, or the daily irrigation extraction from the river system. This day-to-day behaviour was then linked to different environmental, social and economic outcomes. That is, each scenario displayed a set of hydrologic outcomes, which was then translated to a set of triple bottom line outcomes using the environmental, social and economic assessment framework. An example of modelled flow is provided in Figure 2. This hydrograph compares three scenarios, showing modelled flows in the Darling River at Bourke over the same 12-month period. There are a number of conclusions that can be drawn from this 12-month example:  development throughout the river system has significantly reduced flow at this location (a comparison of the dashed and red lines)  Basin Plan activities can restore a portion of flows over most of the year (the blue shaded region)  the majority of the additional volume through modelled environmental water delivery has occurred during existing flow events  recovering water for the environment has also provided another small flushing flow event towards the end of the year Making these types of comparisons over a 114-year period provided a general set of conclusions regarding each modelled Basin Plan scenario. The 114-year modelling period encompassed a wide variety of seasonal and flow conditions (ranging from the wet period in the 1950s to the drought in the 2000s), allowing a detailed analysis of the performance of each modelled scenario under a wide spectrum of possible future flows. For the Northern Basin Review, the social, economic and environmental assessment frameworks used the daily pattern of modelled flows and diversions to measure and compare the outcomes from each scenario. A description of the over-arching triple bottom line assessment tool is given by MDBA (2016b), while companion reports describe the environmental (MDBA 2016c) and socio-economic (MDBA 2016d) bodies of work.

Page 6

Hydrologic Modelling for the Northern Basin Review

Water Act

Basin Plan objectives

Social Economic Hydrology Modelling Environmental Science: UEA • Reviewed socioeconomic approach • Aboriginal community • Reviewed Basin Plan analysis for Basin Plan • Review knowledge survey development modelling • Developed new methods • Projects to fill priority • Community narrative • Developed modelling for analysing impacts of gaps • Consultation report work program to address SDLs at the community • Determine UEAs and site Northern knowledge gaps scale specific ecological targets Basin Review • Scenarios to examine SDL • Use of Census data, • Flow  ecology work program options consultation, detailed assessments • Scenarios to examine economic modelling • Set EWRs options for water recovery patterns & use

Select ESLT options Economic modelling of Hydrology modelling of options for assessment options (range finding)

Social outcomes Economic outcomes Environmental outcomes Consultation and NBAC process advice throughout NBAC and Consultation

Assess outcomes – Triple Bottom Line

Recommend Sustainable Diversion Limit

Figure 1: Overview of the Northern Basin Review work program — the hydrologic modelling component is marked as the dashed outline section

Page 7

Hydrologic Modelling for the Northern Basin Review

30,000 Flow Added With Basin Plan (390GL Recovery)

25,000 Flow Prior to Basin Plan (Baseline Conditions)

Without Development Conditions 20,000

15,000

10,000 Flow at Bourke (ML/day) Bourke at Flow

5,000

0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan

Figure 2: Example 12-month period of modelled daily flows. The red line traces the flow under baseline conditions (i.e. water sharing arrangements prior to the Basin Plan), while the blue region shows the flow added under one of the modelled Basin Plan options (in this case, 390 GL of water recovery). For reference, the dashed line shows the same period under without development conditions.

Page 8

Hydrologic Modelling for the Northern Basin Review

1.3. The Need for Additional Northern Basin Modelling A significant amount of time was required during the Basin Plan development (2009–12) to build and test the hydrological modelling framework. Hence only a limited number of model scenarios testing different SDLs for the Northern Basin were simulated (MDBA 2012a). These scenarios examined the changes in flow that could result from specific water recovery volumes — associated work completed in parallel to the modelling was used to translate these hydrologic changes into environmental outcomes (MDBA 2011). The combined results from the model scenarios and environmental flow work were presented to the Authority during 2011–12. This modelling work informed the Authority’s decision on a northern Basin wide recovery volume of 390 GL. However, it was recognised that this modelling, although the best available at the time, provided only an initial basis for a decision by the Authority. In 2012, only five scenarios had been completed for the Condamine–Balonne, and two for the complete Northern Basin. Further modelling could have provided more clarity regarding the relationship between the SDLs and the resulting flow outcomes. In addition, it was noted that there are ‘levers’ other than straightforward recovery volume that can be used to influence flow outcomes, such as the pattern of water recovery, and strategies underlying the use of this water. The time frame for Basin Plan development did not allow these options to be fully explored. The modelling work program for the Northern Basin Review was built to address these knowledge gaps. Overall, the number of completed scenarios has increased from five to 24 for the Condamine–Balonne; and from two to nine whole-of-north scenarios. This package of scenarios, combined with the companion bodies of work exploring environmental, social and economic outcomes, have provided a broad information base for the Authority to re-examine northern Basin SDLs. In addition to enhancing the information base underlying the SDLs, the modelling explored some aspects of the ongoing implementation of the Basin Plan. Some of the scenarios modelled were designed to examine if the location of future water recovery is likely to have affect system-wide outcomes. Other scenarios examined the impacts in the Barwon–Darling of adopting alternative management strategies for environmental water. Many of the key policy settings underlying these scenarios were based on feedback received from Basin jurisdictions and the community. For the Condamine–Balonne, model scenarios were completed to represent recovery options ranging from 65 GL (an estimate of recovery achieved to date — i.e. no further buyback) up to 176 GL. Further modelling explored whether flow outcomes could be influenced by preferentially recovering entitlements of a certain type or in a specific part of the catchment. The whole-of-north scenarios examined the changes in flow associated with a range total Northern Basin recovery volumes. The options ranged from 278 GL (no further buyback) up to 415 GL. And, as with the Condamine–Balonne, additional modelling explored whether other aspects of Basin Plan implementation could enhance flows through the Barwon–Darling. The modelling has indicated that the spatial pattern of recovery, and the management strategies adopted to direct water into the Barwon–Darling, will be factors in the flows achieved under the Basin Plan.

Page 9

Hydrologic Modelling for the Northern Basin Review

The modelling work has progressively incorporated learnings from all three work programs as the review has advanced. For example, much of the Condamine–Balonne modelling work (Section 6) was completed first, and the social, economic and environmental outcomes from these scenarios were used to guide the subsequent whole-of-north modelling work. Also, the final two whole-of-north scenarios (the ‘refined scenarios’; Section 7.3) were designed based on the conclusions drawn from the prior model scenarios. This learning process is described as part of the scenario descriptions below. Furthermore, the modelling work benefited from the active involvement of partner groups, such as Basin governments (through the Northern Basin Inter-Governmental Working Group) and community representatives (such as the Northern Basin Advisory Committee, Lower Balonne Working Group, and Northern Basin Aboriginal Nations). The modelling work has been presented to many groups throughout the review, allowing many of the modelled policy settings to be tested and refined. The involvement of these groups is acknowledged throughout this document. 1.4. Model Assumptions and Uncertainties Independent reviews confirmed that the modelling platform being used for the Northern Basin Review is fit-for-purpose. However, as with all models, the outputs from the Basin Plan modelling framework are dependent on the underlying assumptions and uncertainties. The modelling uncertainties can be divided into two main themes: those associated with the model calibration, and those associated with the model changes made to represent the Basin Plan. 1.4.1. Inherent Model Uncertainty The models themselves were provided by the Basin States, and have been the subject of development, calibration and testing for up to 40 years. However, in general, model development and calibration has been more robust in the Southern Basin due to regulated nature of these river valleys, their more predictable hydro-geomorphic features, and also as a result of the longer developmental timeframes for these models. In comparison, the Northern Basin is largely unregulated and has less reliable gauge and water loss information on which to base model calibrations. This is particularly true in catchments such as the Condamine–Balonne, Gwydir and Macquarie, which contain large terminal wetlands in which losses vary considerably depending on antecedent moisture conditions. Compounding this issue is the relatively sparse gauge station network across parts of the Northern Basin, allowing only limited information for the calibration of the river system models. The Basin States continue to refine and develop their models to improve their representation of reality. Consideration was given to the updated models developed by jurisdictions since 2012, and some of these models were adopted for the Northern Basin Review. Other models were not adopted due to the following reasons: 1. Insufficient time to ensure that any changes to the Baseline Diversion Limit were verifiable as the new versions of models had not been audited or assessed (e.g. new Barwon Darling model was not received until June 2016) 2. Comparability (important to maintain comparability with the 2012 modelling that informed the Authorities original decision)

Page 10

Hydrologic Modelling for the Northern Basin Review

3. Confidence (the new models were new, but not necessarilly better or final) A full description of the models used for the Northern Basin Review is provided in section 2. 1.4.2. Basin Plan Model Uncertainty The MDBA modified certain aspects of the state-provided models to represent a number of possible permutations of the Basin Plan. Each scenario required two core changes to be made to the existing baseline models: 1. Water Recovery — represent a complete process of water recovery for the environment across the northern Basin 2. Water Use — represent an environmental watering strategy Both of these activities are still ongoing, hence their representation in the models required the MDBA to anticipate future actions. The assumed actions therefore carry their own uncertainties. It is therefore important to recognise that the results contained in this report do not provide a prediction of the outcomes of the Basin Plan. They are instead an indication of the outcomes that could be achieved. The actual outcomes will depend on future policy choices to be made as the Basin Plan continues to be implemented. Most scenarios were completed using a consistent set of assumptions, but explored alternative SDL options. Maintaining unchanged assumptions allowed consistent comparisons of hydrological changes to be made between scenarios, where the changed outcomes can be clearly be linked to the change in SDL. However, additional scenarios were completed in which underlying assumptions were deliberately altered. The purpose of these scenarios was to test the level to which flow outcomes are sensitive to the assumed principles. Hence, while the MDBA acknowledges that the modelled assumptions carry associated uncertainty, the most consequential assumptions have been tested through the modelling framework and the results are outlined in this report. The findings from these scenarios have been presented to the Authority to further inform their decision, but it is also anticipated that these scenarios could help guide the ongoing implementation of the Basin Plan, specifically those aspects related to water recovery and environmental water use. 1.5. Scope of this Report This document describes work conducted under the water recovery modelling theme. The overall outcomes of the Northern Basin Review are provided in the main document (MDBA 2016a). The data provided by the models are hydrological in nature — that is, they are restricted to parameters such as flow, diversions or storage behaviour. The Basin Plan development modelling report (MDBA 2012a) also included environmental flow indicator results with an interpretation of the ecological outcomes for each scenario. This report does not provide the same level of ecological information, and is instead focussed on the hydrological outcomes only. Tables summarising the environmental flow indicator results are included in Appendix A, and these are referenced if they augment the hydrological interpretation of results.

Page 11

Hydrologic Modelling for the Northern Basin Review

A full description of the environmental outcomes associated with each scenario are described by MDBA (2016c). Similarly, the social and economic outcomes associated with each scenario are also described in companion reports by MDBA (2016d).

Page 12

Hydrologic Modelling for the Northern Basin Review

2. Hydrologic Modelling Framework To assist with the development of Basin Plan, the MDBA used the Integrated River System Modelling Framework (Yang 2010) to represent surface water resources. This framework, originally developed for the Murray–Darling Basin Sustainable Yields project (MDBSY; CSIRO 2008), links the 24 river system models that have been developed by Basin State governments, MDBC/MDBA, and Snowy Hydro Limited. The MDBA have since enhanced the capability of the IRSMF to better meet the needs of Basin Plan development. The framework allows policy makers and researchers to better understand the connectivity of the system, and the relationship between upstream changes in the system and downstream impacts. The framework received an independent quality assurance review (Podger et al 2010b) and has since been endorsed in its use for the Northern Basin Review (Bewsher 2016). For each river system, three types of model scenarios were used to inform the Northern Basin Review, each representing a unique set of river system conditions: 1. Without development — all aspects of development have been removed from the model, including infrastructure and consumptive use (i.e. the best available estimate of the natural river system, but without accounting for land use changes and on-farm development). 2. Baseline — the best estimate of water management operations prior to the Basin Plan; this scenario includes all entitlements, water allocation policies, water sharing rules, operating rules and infrastructure such as dams, locks and weirs as of June 2009. 3. Basin Plan — a modification of the baseline scenario to represent a specific permutation of the Basin Plan (such as a chosen water recovery volume, geographic pattern of recovery, and an environmental water management strategy). Baseline and without development models for each catchment were provided by the States, and were placed in the modelling framework to provide ‘whole-of-Basin’ flows under each scenario. In general, these scenarios are rarely updated (i.e. only if new information comes forward), and can be thought of as ‘anchor points’ for the hydrological analysis. The MDBA have completed a large number of Basin Plan model scenarios, each representing different recovery options and water management strategies. The purpose of these scenarios was to explore the changes in flow that could result from a specific combination of Basin Plan settings. These changes, along with the associated environmental, social and economic outcomes, have informed the Authority regarding their re-examination of the Basin Plan settings for the Northern Basin. The individual catchment models are described in detail in the Basin Plan modelling report (MDBA 2012a), and the baseline models were used to estimate the Baseline Diversion Limit (BDL) in each catchment. Barma Water Resources was engaged by the MDBA to undertake a high level independent expert review of these models to assess their representation of baseline diversions (Barma 2012). This review noted that, as is standard practice, Basin States were continuing to update their models to better represent the river system, and that some of these updates were likely to impact the BDL estimate. Model updates are generally made if new calibration data is obtained, or if an aspect of the river system experiences a significant change. Ongoing model updates are an important aspect of

Page 13

Hydrologic Modelling for the Northern Basin Review water resource planning and management. During the course of the Northern Basin Review, Basin States provided the MDBA with updated models for a number of catchments. This collaboration is an important element of ongoing water resource planning, and is gratefully acknowledged by the MDBA. The general approach of the Northern Basin Review was to update the knowledge base from 2012 levels. This principle was applied to the modelling framework, but was adopted with the qualifier that each model had been subjected to sufficient quality assurance and peer review. The models used for the Northern Basin Review modelling are mostly those used for the modelling as part of Basin Plan development, but with updates and improvements to address issues identified at that time. Some of them were minor repairs to errors discovered after the Basin Plan modelling, and some were the incorporation of recommendations from independent audit of Cap models. In general, if any updated model had significantly changed the BDL and had not yet under gone through an independent peer review, it was not adopted for the Northern Basin Review modelling. Work was carried out to see if adoption of the updated NSW models would have significantly changed the achievement of environmental targets, presented in Table 1. Most environmental flow indicators experienced small increases in achievement frequencies, while some of the Barwon–Darling indicators displayed small decreases. The overall changes were all within a few percent. Table 2 shows the differences in terms of diversion between the models given to the MDBA in 2009 and those examined in 2012. It can also be seen that in all cases changes to diversions was less than 10%.

Page 14

Hydrologic Modelling for the Northern Basin Review

Table 1: Comparison of SFI achievement for the baseline scenario (845) and the updated models provided by NSW

Percentage of Years with Event1 Interim Site/Region SFI Baseline Updated Change (845) Models (1030) 4,000 ML/d for 5 days 19 18 -1 Border Rivers 4,000 ML/d for 5 days 38 33 -5 4,000 ML/d for 11 days 16 12 –4 150 ML/d for 45 days 81 84 4 1,000 ML/d for 2 days 85 86 1 45 GL 70 74 4 Gwydir 60 GL 63 63 0 80 GL 46 46 1 150 GL 20 23 3 250 GL 11 11 1 500 ML/d for 75 days 33 36 3 Namoi 1,800 ML/d for 60 days 30 32 3 4,000 ML/d for 45 days 16 18 3 100 GL (total volume) 80 82 2 250 GL (total volume) 35 38 3 Macquarie 400 GL (total volume) 27 27 0 700 GL (total volume) 17 18 2 6,000 ML/d for 14 days 66 64 –2 10,000 ML/d for 14 days 54 56 3 2 × 10,000 ML/d for 20 days 20 18 –3 Bourke 30,000 ML/d for 24 days 20 18 –2 45,000 ML/d for 22 days 17 16 –1 65,000 ML/d for 24 days 10 8 –2 6,000 ML/d for 20 days 58 58 0 Louth 21,000 ML/d for 20 days 32 29 -3 6,000 ML/d for 7 days 42 44 +2 Wilcannia 20,000 ML/d for 7 days 39 38 –1 2,350 GL (total volume) 7 7 0 1Note that these values have been rounded to integers, hence the numbers in the ‘Change’ column are correct but may not match with the integer difference calculated upon examining the numbers in the first two columns.

Page 15

Hydrologic Modelling for the Northern Basin Review

Table 2: Comparison of long-term average diversions for the existing baseline model and the updated models provided by NSW

Long-Term Average Diversions (GL/y) Percentage Catchment Interim Updated Baseline (845) Difference Difference Models (1030) Gwydir 314 320 +6 +2% Namoi 265 274 +9 +3% Macquarie 380 398 +18 +5% Barwon–Darling 198 168* -30 -15% * The WSP W004 Barwon-Darling model is used instead of the former W001, as used in run 1030.

The Northern Basin Review focus was on understanding relative changes in hydrology under various scenarios, and it was concluded that the use of the updated models would not have had a material impact on the findings of the Northern Basin Review modelling. A brief review of the models used in the Northern Basin Review, including a description of the changes made from the versions used in the Basin Plan, is provided below for each model. 2.1. Queensland Models 2.1.1. Paroo Model The Paroo River system was represented with an IQQM model (DLWC 1995). The model used in the Northern Basin Review was the same as used during the development of the Basin Plan, and has been assessed to be representative of baseline conditions (Barma 2012). The model represents the Paroo River from the Yarronvale gauge (424202) to its inflow into Darling River, noting that the Paroo system rarely flows to the Barwon Darling system. The baseline conditions of the Paroo River system (DERM 2006a) are based on the Resource Operation Plan, described in detail by DERM (2006b). The flow duration curve at Wanaaring displays no observable difference between without development and baseline conditions. 2.1.2. Moonie Model The Moonie River system was represented using an IQQM model. The model simulates the Moonie River System from Nindigully (417201) to Gundablouie (417001). The baseline model represents the Resource Operations Plan for the Moonie (DERM 2006d; DERM 2008b). There are no regulated storages, but there are natural pools and water bodies along the length of the Moonie River that are included in the model. The MDBA received the audited cap model from Queensland Department of Science, Information Technology, Innovation and Arts (DSITIA) in 2014. Some changes recommended by the auditors have been reflected in the new model, but the changes were small and had little impact on long-term diversions compared to the Basin Plan development model. The new audited cap model has been used in the Northern Basin Review. The 1.1 GL/y of unallocated water given to the Commonwealth is included in the model as take from system and is part of the BDL.

Page 16

Hydrologic Modelling for the Northern Basin Review

2.1.3. Warrego Model The Warrego IQQM model used for the Northern Basin Review was unchanged from since 2012. The model was assessed to be representative of baseline conditions (Barma 2012). The baseline model represents the Resource Operations Plan of the Warrego system (DERM 2006a, 2006c). The modelled diversions in the NSW part of the catchment are estimates only. The model represents the flow from Augathella gauge (423204) to the confluence with Barwon River. The flows at Fords Bridge gauge (423001) are taken as end of system flows and treated as input into the Barwon Darling model. 2.1.4. Nebine Model The Nebine system is a sub-catchment of the Condamine-Balonne system. The Nebine IQQM model represents headwater inflows at Wallam Creek, Mungallala Creek and the Nebine River and includes inflows from the Warrego River through the Widgeegoara and Noorama Creeks near the NSW border. The model ends at the confluence of Nebine and Warrego inflows. The outflows from the Nebine are inflows into the Lower Balonne system. The Nebine system is unregulated and has no regulated storages, but the model includes nine ‘storages’ to represent natural water bodies (such as waterholes and wetlands). The model represents the Resource Operation Plan of the Nebine (DERM 2006a). In the Northern Basin Review, an updated model incorporating the changes made to meet the requirements of Independent Cap Audit was used. Adoption of these changes increased the long term average total diversions by 0.4 GL/y. The changes comprised corrections to water access parameters at nodes 174 and 156. 2.1.5. Condamine-Balonne Model The Condamine-Balonne system has been represented by four linked models developed by DERM, summarised in Table 3.

Page 17

Hydrologic Modelling for the Northern Basin Review

Table 3: Summary of Condamine–Balonne models Modelling Model Model Description Tool

Condamine headwaters to Cecil Plains Weir (Condamine River) & Lone Pine Gauge (North Condamine River). The model includes Upper IQQM (daily in-stream, unregulated and regulated supply storages. Modelled Condamine time step) water use include high priority supplemented, unsupplemented and overland flow diversion users.

Cecil Plains Weir & Lone Pine Gauge to Beardmore Dam Mid- IQQM (daily headwater gauge. Water use includes high priority supplemented, Condamine time step) medium priority supplemented, unsupplemented and flood plain harvesting users.

Beardmore Dam to Bifurcation 1. The St George water supply scheme is connected to Moolbah and Buckinbah weirs by the Custom capacity Thuraggi Channel. The Beardmore dam and three weirs are St George share model represented by a single storage in the model. The system is (daily time step) operated using a capacity share allocation and accounting scheme and modelled with individual shares in the storage. (BP Modelling report)

Downstream of Jack Taylor Weir to the end of system flows to the IQQM (daily Barwon–Darling & flows to Narran Lakes. Unregulated system, Lower Balonne time step) modelled water use includes unsupplemented, two town water suppliers and overland flow users.

There is some overlap between the St George and Lower Balonne models (specifically the region from downstream of Jack Taylor Weir to Bifurcation 1) — for this overlap region, the data is sourced from the St George model. An updated version of the Condamine–Balonne model was used for the Northern Basin Review. These updates were made after a review of the model for cap auditing purposes, and include the following differences from the models used during Basin Plan development:  Amendments to on-farm storage volumes  Changes to unsupplemented access rules and pump rates  Amendments to operational constraints, such as resetting cap dates in the Mid- Condamine model and revising event management in the St George model.  Removal of domestic and stock demand in the Upper and Mid-Condamine. Adoption of these updates provided minor changes to the long term average diversions (0.4 GL/y reduction). 2.1.6. Border Rivers The Border Rivers and Macintyre Brook systems are modelled separately using two daily time step IQQM models. The Macintyre Brook model simulates the system from Coolmunda Dam to its confluence with Dumaresq River. Coolmunda Dam is the only regulated storage in the model along with two unregulated weirs (Whetstone and Ben Dor). The model operates on an annual accounting system.

Page 18

Hydrologic Modelling for the Northern Basin Review

The Border Rivers model simulates the system from Pike Creek headwater inflows into Glenlyon Dam and the Severn River (NSW) into Pindari Dam. The water use in QLD includes high and medium priority, unsupplemented and town water supplies. In NSW, it includes general security, supplementary and town water supplies. NSW supplementary access is capped at 120 GL/y limit. The model operates as a continuous accounting system. The model corresponds to the Inter-Government Agreement (IGA) between NSW and QLD and is described in detail in DERM (2008a). The models used for the Northern Basin Review were unchanged from those used during Basin Plan development. 2.2. New South Wales Models 2.2.1. Namoi Two separate daily time-step IQQM models were used to simulate the Namoi system, representing the Peel River and the main Namoi catchment. The models have been independently assessed to be representative of baseline conditions (Barma 2012). The Peel model simulates the system from Chaffey Dam headwater inflows to its confluence with Namoi River downstream of the Keepit Dam. The flows at Carol Gap gauge are used as inflows to Namoi model. The Namoi model covers the Manilla River from Split Rock Dam headwater inflows to its confluence with Namoi River, and from the North Cuerindi gauge on Namoi River till its confluence with Barwon River. The flows in Pian Creek at Waminda and Namoi River at Goangra are used as inflows to the Barwon Darling model. The Peel model is operated under an annual accounting system and Namoi model under a continuous accounting system. The development of the Namoi and Peel Cap models is described in detail in DIPNR (2004) and DNR (2006c). The models were reviewed as part of the Cap Auditing and are accredited for Cap implementation (Bewsher 2006b, 2009). The sharing of the supplementary flow events between consumptive and environment use (i.e. the ‘10:90’ rule) is reflected in the model. It is modelled such that between 1st July and 31st October only 10% supplementary water is available for consumptive use and the remaining 90% is for environment. During the rest of the year, supplementary water is available 50:50 percent for consumptive and environmental use. NSW are currently conducting a multi-year trial in which the 50:50 sharing arrangement applies all year (i.e. in which the 10:90 rule no longer applies). The results of this trial are pending, hence MDBA have continued to use the existing water sharing plan arrangements for all NBR modelling. The models used for the Northern Basin Review were unchanged from those used during Basin Plan development. 2.2.2. Gwydir A daily time-step IQQM model was used to represent the regulated parts of the Gwydir catchment. The model encompasses the Gwydir River from Stonybatter gauge to its confluence with the Barwon River, and covers the floodplains of the Mehi River, Mallowa Creek, Moomin Creek and Carole/Gil Gil Creeks. The model connects to the Barwon–Darling model at three end-of-system flow locations: Collymongle gauge (Gwydir River), Collarenebri gauge (Mehi River), and Galloway gauge (Gil Gil Creek).

Page 19

Hydrologic Modelling for the Northern Basin Review

The model represents general security users operating under a continuous accounting system permitted to hold up to 150% of their entitlement in storage, but their usage in any one year is restricted up to 125%. The general security irrigators are also not allowed to take more than 300% in any three years. The model includes an environmental contingency allowance (ECA) entitlement of 45 GL/y of general security water, used to supplement high flow events into key wetlands. The ECA annual entitlement can be carried to a maximum of 200% from one year to another. ECA water is released every year according to the advice of the ECA Operations Advisory Committee. The model represents ECA operations through the following rules  When the total flow past the Yarraman Bridge gauge is between 80GL and 150GL for seven days, immediate releases are made from Copeton Dam to maintain a target flow of 300 ML/d at the Yarraman Bridge gauge.  When the total flow past the Yarraman Bridge gauge is between 150GL and 250GL for seven days, releases are made after 30 days from Copeton Dam to maintain a target flow of 300 ML/d at the Yarraman Bridge gauge.  The targeted flows from Copeton Dam are stopped when the accumulated flow volume during last seven days is greater than 250 GL. The MDBA acknowledge that the management of the ECA has evolved in recent years, and some differences between the on-ground management and the model representation were noted. The model used in the Norther Basin Review is the same as used in the interim benchmark modelling work and is slightly different from the BP model, updated with following two points.  An alteration in the representation of residual catchment inflows.  Repair of errors in the connection between the model and the input data. NSW have recently updated the Gwydir model to include a refined representation of the Gwydir ECA and greater detail regarding flow behaviour within the Gwydir Wetlands. However, while the model shared with the MDBA included changes in estimates of inflows and revised volumes of OFS, it still included an unchanged representation of ECA behaviour and the Gwydir wetlands. There was not sufficient time to incorporate the model in the MDBA framework and fully assess the flow changes resulting from these updates, both within the Gwydir and downstream in the Barwon–Darling. Also, it is unlikely that the use of this interim model would have changed the outcomes of the Northern Basin Review (see the SFI comparison in Table 1). The MDBA have therefore continued to use the Basin Plan development model (albeit updated) which was independently assessed to be fit-for-purpose for exploring SDL options. 2.2.3. Macquarie-Castlereagh A daily time-step IQQM model was used to represent the regulated parts of the Macquarie River, and unregulated parts of the Castlereagh and Bogan Rivers. Flows to the Barwon– Darling are an aggregate of flow at five locations: Carinda gauge (Macquarie River), Carinda gauge (Marthaguy Creek), Coonamble gauge (Castlereagh River), Gongolgon gauge (Bogan River) and Billybingbone gauge (Marra Creek). The development of the model, its calibration and validation, and the set-up for cap implementation is described in detail by DNR (2006a).

Page 20

Hydrologic Modelling for the Northern Basin Review

The system works under an annual accounting system with 100% maximum allocation level. The environmental water for the Macquarie Marshes is a general security allocation and is modelled in two components:  A 64 GL component, released as a fixed pattern every year in June without any carryover provision.  A 96 GL component, used as translucent releases from Burrendong Dam between 15th March to 31st May and between 21st June to 30th November with 100% carryover provisions. These releases attempt to achieve flows between 500 and 4000 ML/d at Marebone gauge. The releases are not made if tributary inflows downstream of Burrendong Dam are above 1000 ML/d. The model used for the Northern Basin Review was unchanged from the Basin Plan development version. NSW Office of Water has recently updated the modelled representation of the Macquarie Marshes to better simulate flow through the distributary network of Macquarie Marshes. There was not sufficient capacity to incorporate the model in the MDBA framework and fully assess the flow changes resulting from these updates, both within the Macquarie and downstream in the Barwon–Darling. Also, the environmental flow results provided by this updated model were near-equivalent to those from the existing model (Table 1). The MDBA have therefore continued to use the Basin Plan development model (albeit updated) which was independently assessed to be fit-for-purpose for exploring SDL options. 2.2.4. Barwon-Darling A daily time-step IQQM model was used to represent the Barwon Darling valley. The model receives tributary inflows from upstream river catchments and routes them through the Barwon Darling system to Menindee. The lower end of model covers the wetlands, lakes and billabongs of the Talyawalka wetland system ceasing upstream of Menindee Lakes. The model has three irrigation class entitlements (denoted as class A, B and C) and also flood plain harvesting diversions. In-stream minimum flows are protected from extraction by specifying flow thresholds for each class licence in different river reaches. The unregulated water use towards the end of the tributary catchments have been represented in Barwon Darling model in the tributary branches before joining the Barwon or Darling Rivers. The high flows bypassing the downstream gauges are reflected in the tributary branches as losses. The models do not include the Interim Unregulated Flow Management Plan for the North West. The model used for the Northern Basin Review was unchanged from the Basin Plan development version. This model represents 2007/08 level of development and incorporates cap accounting rules of July 2007. NSW have recently finalised an updated model representing the water sharing rules of the interim water sharing plan. This model was available relatively late in the Northern Basin Review process, hence there was not sufficient time to incorporate the model in the framework and fully assess the flow and diversion changes resulting from these updates. As a result, the MDBA version of the model does not include the embargo on diversions of Class B and C licences, and also does not include the water sharing rules of interim WSP, such as limiting take to 300% limit in any water year. Although the updated model was not used as part of the Northern Basin Review, a comparison of two versions of the model indicated only

Page 21

Hydrologic Modelling for the Northern Basin Review small differences in long term average flow values and environmental flows (see Table 1) along the system and confirmed that the use of the water sharing plan model would not have impacted the key findings from the Northern Basin Review modelling. Community consultation conducted as part of the Northern Basin Review has emphasised the social and cultural importance of low flows through the Barwon–Darling. Work by the MDBA suggests that rule changes in recent years may have reduced the protection of low flows, but this reduction will not be reflected in the Northern Basin Review modelling results. The Authority have therefore recommended (as part of the ‘toolkit’) improvements to state water management arrangements to safeguard low flows across the North (MDBA 2016). 2.3. Without Development Scenario The without development scenario represents near natural conditions of a river system. It is based on the baseline scenario and is generated by removing all the dams, irrigation and environmental works and infrastructure from the baseline model. All the consumptive users (such as irrigation, town water supply and industrial water use) and the flow governing rules (such as channel capacities, minimum flow requirements, maximum flow constraints etc.) are also removed from the model. This scenario does not necessarily represent the pre-European conditions, as the inflow data have not been adjusted for land use changes. Also, the impact on flows due to levee construction and in-channel structures are also not considered. However, this scenario is the best available representation of the natural conditions of the river system. 2.4. Baseline Scenario The baseline scenario represents the water sharing arrangements that were in-place in June 2009. This includes the entitlements, water allocation policies, water sharing rules, operating rules and infrastructures such as dams, locks and weirs existing as of 2009. The level of development is as per the Murray-Darling Basin Cap for all States, unless current water sharing arrangements have a usage level lower than the Cap level (e.g. NSW water Sharing Plans). 2.5. Basin Plan Scenarios A number of ‘Basin Plan scenarios’ were completed for the Northern Basin Review. The purpose of these scenarios was to explore the changes in flow that can be achieved through the recovery and use of water for the environmental under the Basin Plan. Each scenario was based on a specific set of input parameters, and was therefore designed to explore a specific permutation of the Basin Plan. Many of the scenarios were built to examine an SDL (e.g. 320 GL recovery across the Northern Basin), while other scenarios were designed to explore the changes associated with another aspect of Basin Plan implementation (such as alternative options for directing water to the Barwon–Darling system). The overall modelling approach is described in Section 4. The starting point for all Basin Plan scenarios was the baseline scenario — this model was modified to represent the two key hydrological changes prompted by the Basin Plan:  Environmental water recovery — the model is altered to represent the water recovery process; a set proportion of entitlements are recovered from each catchment and placed in a proxy environmental water account.

Page 22

Hydrologic Modelling for the Northern Basin Review

 Environmental water management — water from the proxy environmental account provides an increase in flows through the river system, achieving an enhanced set of environmental outcomes. Making these changes requires an anticipation of the future. In practice, water recovery is ongoing, and environmental water holders are still improving the effectiveness of their portfolio to achieve environmental outcomes. For this reason, the MDBA have made a set of informed estimates (i.e. modelling assumptions) about the end-point of these ongoing processes. These assumptions are further described in Section 5.

Page 23

Hydrologic Modelling for the Northern Basin Review

3. Flow and Connectivity Through the Northern Basin 3.1. Climate, Rainfall, Flow and River Regulation The Northern Basin (Figure 3) encompasses a number of catchments over a large area, each with a unique set of natural characteristics such as climate, inflows and geomorphology. These characteristics have influenced the development of water resource management, such that each catchment has flow regulation levels (through public storages), water sharing arrangements, and a proportion of take from the river system particular to that catchment. Queensland catchments receive most of their rainfall in summer months. There is a strong east- to-west rainfall gradient with the majority of rainfall occurring in the eastern, upland regions. Year-to-year rainfall variation is high, hence river flows are extremely variable with dry-to- median periods interspersed with very wet years. Large rainfall and flow events usually occur as a result of deep subtropical depressions.

Figure 3: Schematic diagram of the main rivers in the Northern Basin Most Queensland catchments are near-unregulated. The Condamine–Balonne and Border Rivers both contain public storages but, compared to those further south, they are relatively small and provide only a limited ability to regulate flow. Most consumptive use is therefore

Page 24

Hydrologic Modelling for the Northern Basin Review associated with unregulated licence conditions, which allow a user to extract water from the river once flow reaches a certain height. Moving south, the rainfall become less summer-dominant. Over the long-term, the Macquarie– Castlereagh catchment displays little seasonal variation in rainfall. Furthermore, rainfall in these catchments displays less year-to-year variability (i.e. greater reliability). Most of the rain (and hence most of the river inflows) occur in the headwaters located eastwards towards the Great Dividing Range. New South Wales catchments display a higher level of flow regulation compared to those in Queensland. Figure 4 compares the relative regulating capacity of the major public storages throughout the Northern Basin. In this graph, regulating capacity is defined as the ratio of active storage capacity to average annual inflows. For example, a ratio of 200% indicates that a storage can hold twice the average annual inflow, hence a higher value indicates a greater ability to store and regulate upstream inflows. Large public storages located in the upper reaches of each catchment collect water throughout the year, most of which is released when required for downstream irrigation purposes. These storages also provide water supplies for towns along the river and, in wet times, can be used to mitigate floods. Figure 4 does not include private storage capacity. Some users in the unregulated catchments (particularly the Barwon–Darling and Condamine–Balonne) own off-river storages with relatively large capacities. These allow extracted water to be stored for use later in the season. Off-river storages do not provide the same flow regulating capacity as on-river public storages, but their development allows for greater irrigation and has therefore had an impact on flows. As an example, private storage capacity in the Condamine–Balonne (estimated to be 1,582 GL; Webb, McKeown and Associates Ltd 2007) is significantly greater than that provided by on-river public storages (around 230 GL).

Figure 4: Ratio of active storage capacity to average annual inflow for the major public storages throughout the Northern Basin The Northern Basin is drained by the large, semi-arid Barwon–Darling River. This river experiences one of the most variable flow regimes in the world, with low flows punctuated by episodic flood events that inundate extensive areas of floodplain. Due to the generally hot and

Page 25

Hydrologic Modelling for the Northern Basin Review dry climate, and the largely flat topography of the catchment, the Barwon–Darling receives very little in-catchment runoff. Flows in the river are therefore almost entirely reliant on inflows from tributary catchments. The Northern Basin is generally considered to terminate at Menindee Lakes, a natural lake system that was developed as a water storage in the 1960s to enhance flow regulation capacity through the Lower Darling and Murray Rivers. 3.2. Connectivity Longitudinal connectivity (or simply connectivity2) refers to the capacity to which each catchment can provide flow into the Barwon–Darling River. Some catchments, such as the Namoi and Border Rivers, display a higher level of connectivity, such that flows are transmitted from these catchments into the Barwon–Darling with a relatively high level of efficiency. Other catchments, such as the Condamine–Balonne, Gwydir and Macquarie–Castlereagh, contain large floodplain and wetland systems that absorb significant volumes of water just upstream from the junction with the Barwon–Darling River. Floodplains act as natural flow attenuating structures in the landscape — a high flow event entering these systems will be slowed and its peak will be depressed. These catchments contribute flow into the Barwon–Darling, but these events are usually a long, low flow that lasts weeks or months. These catchments also tend to contribute large volumes downstream only when the floodplains and wetlands have absorbed significant volumes of water (i.e. they are pre-wetted). At the low end of the connectivity scale, the Paroo River is essentially a disconnected system that only contributes water downstream during extreme flooding events that occur once or twice a century. 3.3. Quantifying Northern Basin Connectivity The Basin Plan seeks to achieve whole-of-basin outcomes. It recognises that enhancing flow through the Barwon–Darling River will depend on actions across the entire Northern Basin, including water recovery and environmental water delivery from upstream catchments, and additional actions in the Barwon–Darling itself (such as water recovery and/or complementary measures). To inform the development of the environmental watering strategy represented in the modelling framework, the MDBA completed an initial study to quantify catchment connectivity throughout the Northern Basin. Of particular interest were the following questions: 1. What was the pre-development level of connectivity of each catchment with the Barwon– Darling River? 2. How has development of the river changed connectivity? 3. What are the detailed hydrological characteristics of flows provided downstream by each catchment?

2 This report will use ‘connectivity’ to refer to the downstream longitudinal aspect of river connectivity. ‘Lateral connectivity’, referring to the level of connection between a river and its adjacent floodplains, wetlands and marshes, is also an important hydro-geomorphological and ecological concept that is discussed in the environmental outcomes report (MDBA 2016c).

Page 26

Hydrologic Modelling for the Northern Basin Review

A useful first test was to examine the long-term average volumetric contribution of each catchment to the Barwon–Darling. This was examined for both the without development and baseline models in Figure 5. The without development model, although not a perfect representation of the natural river system, indicates how flows would appear without extraction or flow regulation by structures such as dams and weirs. A comparison to the baseline model shows changes that have occurred as a result of human development. Figure 5 indicates that the vast majority of water flowing into the Barwon–Darling River is sourced from the Condamine–Balonne, Border Rivers, Gwydir, Namoi, and Macquarie– Castlereagh catchments. A comparison of the without development (blue) and baseline (red) results shows the volume of water flowing out of each catchment has reduced as a result of water resource development.

Figure 5: End of system flows for each catchment in the Northern Basin for both without development and baseline (pre-Basin Plan conditions) models The relative efficiency of each catchment in passing water downstream can be further explored through the inflow-to-end-of-system transmission. This transmission factor is a measure of the proportion of inflowing water reaching the end of system for each catchment, and is shown for each northern catchment in Figure 6. A transmission value of 1 indicates that 100% of inflows reach the end of system (the theoretical ‘lossless system’), whereas a value of zero indicates that none of the inflows reach the end of system. This calculation was completed for both the without development and baseline model outputs. The without development transmission ratios (blue bars) reflect the proportion of water ‘lost’ to natural effects such as evaporation and exchange with groundwater systems. It can be seen that some systems (notably the Moonie) pass a relatively high proportion of inflows downstream, whereas most of the water in a system such as the Warrego is lost naturally.

Page 27

Hydrologic Modelling for the Northern Basin Review

A comparison to the baseline results (red bars) shows the reduction in connectivity that has occurred in each catchment as a result of water resource development. There are two overall process that underlie this reduction. Large dams absorb and store inflows to be released when required as in-channel flows, while extraction and utilisation results in less water passing downstream. The most significant reductions (i.e. more than 50%) have occurred in the Condamine–Balonne and the Gwydir systems. The Condamine–Balonne has only a small level of flow regulation, hence extraction of water from the river is the primary driver of this result. In the Gwydir, public storage and consumptive use are both contributing factors. A comparable analysis was completed by CSIRO as part of their Murray–Darling Basin Sustainable Yields project (CSIRO 2008), with similar conclusions.

Figure 6: Proportion of inflows reaching the end of system for each catchment in the Northern Basin for both without development and baseline (pre-Basin Plan conditions) models This analysis addresses the first two questions listed above, but it provides only a long-term volumetric perspective of connectivity. It does not consider the types of tributary flows from each catchment. Flows through the Barwon–Darling are complex. With multiple tributaries experiencing different climates, it is difficult to define a standard event through this system. Examples of this complexity are shown in Figure 7. The upper panel shows a summer flow at Bourke (dotted line) that originated mostly in Queensland catchments. The lower panel displays two winter-spring events that were mostly sourced from New South Wales catchments. The hydrographic shape of both events appear near-equivalent when measured at Bourke, but they represent the outcome of two different types of weather event.

Page 28

Hydrologic Modelling for the Northern Basin Review

Figure 7: Examples of a Barwon–Darling event sourced primarily from Queensland catchments (upper panel) and NSW catchments (lower panel). The dotted line traces the flow at Bourke and the coloured lines trace the end of system flows from each catchment.

Page 29

Hydrologic Modelling for the Northern Basin Review

Figure 8: Schematic representation of the average tributary contribution to a flow event at Bourke in the range 15,000 – 20,000 ML/d

Page 30

Hydrologic Modelling for the Northern Basin Review

To examine this detailed aspect of connectivity, the 114-year modelled hydrograph at Bourke was divided into a series of distinct flow events, and each event was matched with the contributing flow event from the tributary catchments. From this, an average tributary flow could be derived for each catchment. An example of the output from this process is shown in Figure 8 for flow events in the range 15,000 – 20,000 ML/d at Bourke. This average event does not represent a proposed management strategy for future flows. It is unlikely that this type of event would occur in practice, as it would require a contrived set of circumstances for all catchments to provide an average contribution at the same time. The advantage of this approach is that it removes climate from consideration and instead allows an objective comparison of the different hydrologic characteristics of inflows from each tributary to the Barwon–Darling. It can be seen that most of the peak of the average Bourke event has been provided by the Namoi and Border Rivers catchments. This emphasises the relatively high level of downstream connectivity displayed by these catchments. Relatively short-duration high- peaked flows through these systems are passed downstream with an essentially unchanged shape. The Moonie River has a similar level of connectivity, noting that this is a smaller catchment with less water flowing through the system. In contrast, the contributions of the Macquarie and Condamine–Balonne catchments have a lower peak flow but a longer duration. This reflects the attenuating nature of the floodplains towards the end of these systems. A similar effect is seen in the Gwydir contribution, again noting that this is a smaller catchment with lower inflow volumes. The inherent connectivity characteristics of each catchment informed the environmental water strategy adopted for the modelling (described in section 5).

Page 31

Hydrologic Modelling for the Northern Basin Review

4. Modelling Approach The modelling approach for the Northern Basin Review was developed to address gaps in knowledge from Basin Plan development. It was acknowledged at the time (MDBA 2012a) that there were two main areas that could be substantially improved through additional refined modelling:  Recovery volume — the number of scenarios completed for the Condamine–Balonne (five) and whole-of-north (two) did not provide enough resolution to accurately measure the relationship between recovery volume and outcomes.  Recovery pattern & utilisation — the scenarios did not measure the sensitivity of outcomes to the pattern of recovery (in terms of both location and entitlement type) or the adopted strategy for the use of environmental water. The modelling work program was therefore designed to be a stochastic process exploring many aspects of recovery, requiring a relatively large number of scenarios. The modelling work program was split into two phases. Phase I included a number of ‘range- scoping’ scenarios testing water recovery parameters over a wide range. The learnings from this phase informed the scenarios to be tested in Phase II. These Phase II scenarios also included new flow indicators informed by the latest environmental science information obtained for the review. 4.1. Condamine–Balonne Modelling The Condamine–Balonne region lies mainly in southern Queensland and extends about 100 km south-west into New South Wales. A schematic map of the region, including structural features and flow constraints, is given in Figure 9. From a water management perspective, the Condamine–Balonne is pre-dominantly an unregulated region with a relatively high level of extraction. There is some regulated take from the system, the majority supplied by Beardmore Dam and Jack Taylor Weir to the St George Water Supply Scheme. But most irrigation production relies on diverting unregulated flows into large privately-owned off-stream storages, particularly downstream of St George. Within the pool of unregulated use, there is a wide spectrum of entitlement conditions, each associated with certain types of flow rates (i.e. flow heights). Under baseline conditions (i.e. prior to the Basin Plan), long-term average watercourse diversions under full uptake of entitlements from the Condamine–Balonne were 713 GL/y (42% of inflows this catchment).

Page 32

Hydrologic Modelling for the Northern Basin Review

Figure 9: Schematic diagram for the Condamine–Balonne; the Lower Balonne Floodplain and Narran Lakes areas are outlined in green, and the locations to monitor modelled environmental flows are marked in light blue During the development of the Basin Plan (i.e. prior to 2012), the MDBA completed five model scenarios exploring recovery volumes in the Condamine–Balonne of 60, 100, 130, 150 & 203 GL (MDBA 2012a). The purpose of these scenarios was to determine the relationship between recovery volume and flow outcomes — the results from these scenarios formed part of the evidence base on which the Authority selected a local recovery volume of 100 GL for this system. However, it was recognised at the time that, “the complexity of entitlement types, in combination with the generally unregulated nature of this system, suggests that the location and type of purchased entitlement will have a significant influence on the ability to meet the desired flow outcomes” (MDBA 2012a). That is, the evidence suggested that the flow outcomes in the Condamine–Balonne would be dependent not only on the volume of recovery, but also on the pattern of recovery. Time constraints during the development of the Basin Plan did not allow a full exploration of these variables. The model scenarios completed prior to 2012 therefore concentrated water recovery in the lower parts of the catchment (primarily in the Lower Balonne), under the assumption that this strategy would provide the best ‘environmental outcomes-per-GL’ return. However, adopting a highly targeted strategy in practice would impose significant limits on the water recovery process, effectively excluding much of the consumptive pool. Furthermore, it was noted that

Page 33

Hydrologic Modelling for the Northern Basin Review a targeted strategy would likely focus the potential impacts of water recovery on a specific part of the catchment. At the time, there was reasonable logic underlying the targeted strategy, however these assumptions had not been tested through the modelling platform. To address these gaps in knowledge, a large number of model scenarios were completed as part of the Northern Basin Review. For the purpose of this study, the Condamine–Balonne region was divided into six sub-regions, displayed graphically in Figure 10.

Figure 10: Schematic map of the Condamine–Balonne including the six sub-regions considered for the spatial modelling work (from north to south, the six sub-regions are ‘Streams of the main river stem’, ‘Upstream Beardmore’, ‘St George’, ‘Jack Taylor Weir to Bifurcation 1’, ‘Narran System’, ‘Lower Balonne System’) Each water recovery parameter was explored individually. For example, to determine the dependence of flow changes on the spatial distribution of recovered water, a set of scenarios were completed in which all other parameters (such as total recovery volume for the Condamine–Balonne, and types of entitlements recovery) were held constant, and the only variation between scenarios was the spatial distribution. This process allowed the influence of each parameter to be individually explored without introducing ‘noise’ by varying other parameters. A schematic of the completed model scenarios against water recovery is shown in Figure 11. As a result, the Condamine–Balonne modelling work can be thought of as comprising several distinct packages of model scenarios, each focussed on a single parameter and providing a coherent set of conclusions regarding this parameter. Four packages of model scenario were completed — three exploring different aspects of the water recovery pattern, and a fourth exploring the water recovery volume. The model scenario packages are: Page 34

Hydrologic Modelling for the Northern Basin Review

1. Entitlement location — these scenarios explored the level to which environmental outcomes could be influenced by the geographic location of the recovered water; each scenario represented an alternative split of the recovered water between four sub-regions: upstream of Beardmore Dam; Jack Taylor Weir to Bifurcation 1; Bifurcation 1 to Bifurcation 2 & the Narran River; the Lower Balonne distributary network. 2. Entitlement type — these scenarios represented a wide range of possible recovery portfolios downstream of Jack Taylor Weir, each with a distinct split between water harvester and overland flow entitlements. 3. Horizontal slicing — scenarios completed in collaboration with Qld officials; horizontal slicing is essentially a specialised form of recovery by entitlement type, but instead targets specific flow windows through the recovery of partial entitlements. 4. Volume of recovery — these ‘volume-dependence’ scenarios formed the primary set of information provided to the Authority regarding the SDL, and included a set of assumed settings based on the findings from the water recovery pattern scenarios. Most of the volume-dependence scenarios are nested — that is, they used the water recovery achieved to date (65 GL) as a starting point, and built the desired volume (such as 90, 100, or 150 GL) using a consistent set of principles derived from the water recovery pattern scenarios. The recovery patterns included in the 24 completed model scenarios are presented in Table 4 and shown graphically in Figure 11. In many cases, the setup of each scenario allowed it to be used for multiple purposes. For instance, a ‘spatial sensitivity’ scenario could also be used to investigate the sensitivity of flows to entitlement mix. The scenarios used to explore each parameter are listed in Section 6.

Page 35

Hydrologic Modelling for the Northern Basin Review

Table 4: Breakdown of baseline diversions and water recovery by sub-catchment for all Basin Plan Condamine–Balonne model scenarios; the grey row indicates the sub-region that was not considered for recovery as part of this study. Model scenarios are divided into parameter of interest (e.g. ‘Initial SDL Scenarios’) and by model run number (e.g. 980). Recovery Volume by Model Scenario (GL) Baseline Horizontal Slicing Refined SDL Scenarios Diversions Spatial Recovery Entitlement Type (collaborative Initial SDL Scenarios (scenarios 1103 & 1115 include supplemented low flow Entitlement Type (GL/y) Sensitivity Sensitivity scenarios with Sub-Region releases from Beardmore Dam) Recovered in Model Qld)

1089 1089 &

1023 1032 1022 1009 1010 1037 1040 1046 1047 1048 1041 1043 1044 1113 1112 1114 1111 1110 1108 1109 1103 1115

845 980

Upstream Beardmore — 189.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Dam (off main stem) Upstream Beardmore Unregulated 97.0 0 0 0 0 10 20 0 30 0 0 0 0 0 0 7 15 10 19 10 10 15 10 10 Dam (on main stem) Supplemented Medium St George 78.6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 4 4 4 4 4 4 Priority Water harvester 123.9 15 9 9 9 9 10 15 45 15 34 20 52 34 47 15 18 15 24 25 25 30 15 24 (i.e. unsupplemented) JTW to B1 Overland flow 18.0 0 1 1 1 1 1 0 0 18 0 14 0 17 0 0 0 0 0 13 13 15 5 0 (i.e. floodplain harvester) Water harvester 27.2 9 11 16 29 19 15 15 8 11 24 14 13 13 13 7 15 20 21 11 15 17 11 21 (i.e. unsupplemented) Narran system Overland flow 12.8 1 12 12 12 11 7 5 3 11 0 10 2 2 2 0 2 3 2 10 10 13 10 2 (i.e. floodplain harvester) Water harvester 114.6 9 35 34 47 46 46 70 30 31 55 40 12 12 12 21 21 12 26 45 48 52 25 26 (i.e. unsupplemented) Lower Balonne Overland flow 52.1 12 33 41 43 44 44 34 22 53 27 42 17 17 17 16 15 41 20 25 25 30 20 28 (i.e. floodplain harvester)

TOTAL 714.5 47 100 113 141 141 142 140 139 140 140 140 97 96 92 65 90 101 115 143 150 176 101 115

Page 36

Hydrologic Modelling for the Northern Basin Review

U/S Beardmore St George Supplemented St George to B1 — Water Harvesters St George to B1 — OLF & FPH Narran — Water Harvesters Narran — OLF & FPH LBF — Water Harvesters LBF — OLF & FPH

1109 1108 1110 1089 1040 1010 1009 1022 1047 1048 1046 1037 1111 1115

ModelScenario 1032 1114 1103 1023 1041 1043 1044 1112 1045 1113 980

0 20 40 60 80 100 120 140 160 180 200 Modelled Water Recovery (GL)

Figure 11: Condamine–Balonne model scenarios completed as part of the Northern Basin Review. The aggregated bar shows the total water recovered in the catchment, sub-divided by colour to represent sub-region and entitlement type recovered.

Page 37

Hydrologic Modelling for the Northern Basin Review

4.2. Whole-of-north Modelling The Barwon–Darling is a large semi-arid river system draining the northern section of the Murray–Darling Basin. A schematic map of the region is given in Figure 12. As described in section 3, each tributary catchment has a unique set of characteristics such as climate, inflows, geomorphology, downstream connectivity, level of regulation, and proportion of take from the river system. As a result, the types of flow contributed downstream by each catchment are also unique. Development of the water resource, through the construction of on-river infrastructure such as dams, and diversion from the river, has changed the contribution of each tributary system. Under baseline conditions (i.e. prior to the Basin Plan), long-term average watercourse diversions from the Northern Basin were 2,541GL/y (23% of total inflows). There are four primary levers that can be used to influence flow through the Barwon–Darling:  the volume of water recovery across the northern Basin;  the pattern of water recovery (i.e. location and entitlement type) across the northern Basin;  the management of water into the Barwon–Darling; and,  the management of water through the Barwon–Darling. During the development of the Basin Plan (2009–12), only the first lever was explored. Two whole-of-north scenarios were completed as part of the development of the Basin Plan, representing northern Basin recovery volumes of 390 and 440 GL. Both of these scenarios were identical in recoveries from each catchment, except for a difference of 50 GL from the Condamine–Balonne. Therefore these scenarios only provided an initial set of information to inform the SDLs recommended for the northern Basin in 2012. Also, they did not reveal the extent to which flows could be influenced by altering the recovery pattern between catchments, or the types of entitlements recovered within the catchments, or by the assumed environmental water management strategy. Nine key whole-of-north scenarios were completed to inform the Northern Basin Review. A schematic of these scenarios is shown in Figure 133. This schematic traces the recovery volume modelled in each scenario, but it also shows the core assumptions included in each scenario. An examination of these assumptions demonstrates that three of the four levers identified above were modelled for the Northern Basin Review. The fourth lever (management of water through the Barwon–Darling) was examined through a separate analysis of model output data described in section 8. The whole-of-north modelling work was divided into three distinct sets of model scenarios. As a general guide, the environmental, social and economic findings from the first two sets of

3 This figure refers to the recovery approach adopted for the Condamine–Balonne. ‘Non-targeted’ refers to near-pro rata approach downstream of Beardmore Dam, similar to that adopted in the Basin Plan modelling prior to 2012. ‘Targeted’ refers to an approach in which a specific geographic distribution and entitlement type has been adopted for specific aims (e.g. to maximise Lower Balonne Floodplain flows) without regard for existing recovery (i.e. it assumes that the Commonwealth can redistribute their portfolio at will). ‘Semi-targeted’ represents a targeted approach that recognises recovery that has already been achieved (i.e. the Commonwealth builds on their existing portfolio).

Page 38

Hydrologic Modelling for the Northern Basin Review modelled scenarios were provided to the Authority, which then narrowed the range of options to be modelled in the third ‘refined scenarios’ set. The scenario sets are summarised below. 1. SDL Range-Finding (Section 7.1) These scenarios quantified the overall relationship between recovery volume and flow, and were analysed with the updated environmental, social and economic assessment frameworks under the Northern Basin Review. 2. Alternative Policy Settings (Section 7.2) Three scenarios were completed to test the effects of altering one of the underlying policy settings — specifically, the assumed distribution of future water recovery, and the assumed management pattern for delivering water to the Barwon–Darling (i.e. a comparison of environmental flow management Strategies 1 and 2). The scenarios indicated that both recovery pattern and the management strategy can be used to influence flow outcomes. 3. Refined Scenarios (Section 7.3) The final two scenarios were built using the environmental, social and economic results from previous model scenarios (SDL range-finding and alternative policy settings). These scenarios encompassed the narrowed range of options under consideration by the Authority in the final stages of the Northern Basin Review. The first set of scenarios (marked in blue in Figure 13), explored a recovery volume range between 278 GL (an estimate of existing recovery) and 415 GL. All of these scenarios were based on the same underlying principles to ensure that changes in flow outcomes could be confidently associated with changes in recovery volume. The second set of scenarios are marked in orange (altered spatial distribution of recovery) and yellow (alternative flow management practices). The refined scenarios are marked green. The distribution of water recovery by catchment for each model scenario is listed in Table 5, and is shown graphically in Figure 14. Some of the scenarios included a targeted recovery of certain types of entitlement in the Condamine–Balonne. The specific targeting aspects were based on the environmental, social and economic findings from the exploratory scenarios described in Section 6, and are described further in that section.

Page 39

Hydrologic Modelling for the Northern Basin Review

CONDAMINE-BALONNE Weir

MOONIE BORDER RIVERS Macintyre WARREGO Boomi

PAROO Gwydir GWYDIR Barwon Mehi Bourke Collarenebri River Pian Brewarrina Walgett Namoi NAMOI/PEEL Louth Castlereagh Bogan Macquarie Darling River Tilpa MACQUARIE- CASTLEREAGH

Wilcannia Poopelloe Lake

Menindee Lakes

To Darling River (Lower Darling Region)

Figure 12: Schematic map of the Barwon–Darling system and its main tributaries

Page 40

Hydrologic Modelling for the Northern Basin Review CORE MODELLING ASSUMPTIONS

FLOW MANAGEMENT Coordinated releases No coordination Coordinated releases Coordinated releases Coordinated releases

SHARED RECOVERY Existing + Default Existing + Default Geographically Targeted Default Geographically Targeted

C–B ENTITLEMENT TYPE Non-Targeted Non-Targeted Targeted Non-Targeted Semi-Targeted

450

415 GL

400 Basin Plan W/O Flow Coord 390 GL 390 GL

350 C–B 100 GL Targeted Refined Recovery 350 GL 345 GL Refined Recovery 320 GL 320 GL Default

321 GL NORTH RECOVERY VOLUME (GL) VOLUME RECOVERY NORTH

- 300 OF - Existing Recovery 278 GL

WHOLE 250

Figure 13: Schematic map of the whole-of-north model scenarios completed for the Northern Basin Review. The vertical axis traces the Northern Basin recovery volume, while the horizontal axis traces the core modelling assumptions included in each scenario. A definition of the terms (e.g. ‘targeted’, ‘semi-targeted’) is provided in the text.

Page 41

Hydrologic Modelling for the Northern Basin Review

Table 5: Catchment water recovery distributions represented in the Northern Basin Review model scenarios; grey cells indicate catchment recoveries that are unchanged from current recovery volumes and were not altered through the modelling work program.

Modelled Recovery (GL) (Model Run Number)

SDL Resource Unit Scenario D Scenario E Scenario G Scenario J Scenario I Scenario C Scenarios B & H Scenario A 278 GL 320 GL 320 GL PR 321 GL 345 GL 350 GL 390 GL 415 GL (1113) (1112) (1111) (1115) (1103) (1114) (1089 & 1110) (1108) Paroo 0 0 0 0 0 0 0 0

Warrego 8 8 8 8 8 8 8 8 Nebine 1 1 1 1 1 1 1 1 Moonie 2 2 2 2 4.5 2 2 2 Condamine-Balonne 65 90 115 115 100 100 142 150 QUEENSLAND Queensland Border Rivers 15 21 21 21 35 25 23 25 Queensland Total 91 122 147 147 148 136 176 186

Intersecting Streams 8 8 0 8 8 8 8 8 Gwydir 48 48 51 47 47 56 56 59 NSW Border Rivers 3 7 13 7 7 16 16 18 Namoi 13 20 20 20 24 24 24 28 Macquarie-Castlereagh 83 83 77 55 74 83 83 88 Barwon-Darling 31 31 12 36 36 28 28 28

NEW WALES SOUTH NSW Total 187 198 173 173 196 214 214 229 Whole of North Total 278 320 320 321 345 350 390 415

Page 42

Hydrologic Modelling for the Northern Basin Review

Paroo, Warrego, Nebine, Moonie, Intersecting Streams Condamine-Balonne Queensland Border Rivers NSW Border Rivers Gwydir Namoi Macquarie-Castlereagh Barwon-Darling

415 GL

390 GL

350 GL

345 GL

321 GL

320 GL PR

320 GL

278 GL

0 50 100 150 200 250 300 350 400 450 500 Northern Basn Modelled Water Recovery (GL)

Figure 14: Water recovery distributions represented in each of the NBR whole-of-north model scenarios (note that scenarios B and H both represented the same 390 GL recovery pattern)

Page 43

Hydrologic Modelling for the Northern Basin Review

5. Modelling Methodology As described above, the States provided two versions of the model for each catchment: a without development model (a representation of flows without consumptive use and associated large-scale river regulation works) and a baseline model (representing water sharing arrangements prior to the Basin Plan). Each Basin Plan scenario was built by modifying the existing baseline scenario to represent a fully implemented Basin Plan with a specific group of settings — that is, each scenario was different depending on the assumed settings. The core Basin Plan setting under investigation was the SDL, but the MDBA have also explored other aspects of Basin Plan implementation, such as the future apportionment of the shared recovery component, or the future management of water recovered for the environment. A full description of the Basin Plan modelling methodology is given by MDBA (2012a), but the approach requires two core changes to be made to the existing baseline scenario: 1. A representation of water recovery 2. An assumed environmental water management strategy

The same overall approach was adopted for the NBR scenarios, although some of the detailed aspects have been refined since 2012. The principles and practices underlying each of these model changes are described below. In determining key policy settings which underpin the modelling methodology, the MDBA has taken into consideration input and feedback received from relevant Basin jurisdictions and community feedback (through representative groups such as NBAC). 5.1. Representing Water Recovery The representation of water recovery included two steps. First, the shared recovery component was distributed amongst catchments, defining a total SDL for each catchment. And second, the desired SDL was achieved by implementing a water recovery process in the models. 5.1.1. Apportionment of the Shared Component Under its current settings, the Basin Plan requires 390 GL to be recovered across the Northern Basin. This volume includes local reductions from each catchment (totalling 247 GL), plus an additional 143 GL shared recovery to achieve environmental outcomes in the Barwon–Darling. The Basin Plan includes a default method for distributing this shared component (based on the relative baseline diversions of each catchment), however it also provides Basin governments flexibility to choose an alternative distribution. For the purposes of modelling, existing recovery was assumed to include the recovery that had already been achieved under the Basin Plan (267 GL as of December 2015), plus an additional 11 GL of recovery that was estimated to be achievable through future investment in infrastructure upgrades — a total of 278 GL. Assuming a default distribution of the shared component, existing recovery in some catchments (Table 6) has already exceeded the aggregate of the local and shared recovery. Rather than represent a purely default distribution of the shared component, the Northern Basin Review modelling includes recovery achieved to date, and uses a set of principles to distribute the remaining water recovery. The overarching assumption for the modelling was Page 44

Hydrologic Modelling for the Northern Basin Review that any future recovery would build on the existing portfolio of environmental water held by the Commonwealth. That is, ‘existing recovery’ formed the starting point for the modelled scenarios. The principles listed below provide an ‘existing + default’ method for distributed the shared recovery. Most NBR scenarios were based on these principles (some of the later scenarios represented a refined recovery pattern informed by earlier findings, or represented a near- pure default distribution). 1. Existing water recovery formed the starting point (‘existing recovery’ includes water recovered under the Basin Plan to date, plus an estimate of the recovery that could be achieved in the future through infrastructure enhancements; see Table 6); 2. Each catchment was required to meet its local recovery volume 3. Due to one or both of the following, no further contribution to the shared reduction amount was sought from the Paroo, Warrego, Nebine, Moonie and Intersecting Streams systems:  a low level of hydrologic connection between the river network and the Barwon- Darling system and/or;  has met both their in-valley and default contribution to the northern Basin Shared reduction amount. 4. The total SDL for the remaining catchments was the maximum of either:  existing recovery (Table 6), and;  the combination of the local recovery plus the default shared recovery target.

Page 45

Hydrologic Modelling for the Northern Basin Review

Table 6: In-valley and existing recovery volumes for Northern catchments

In-Valley Existing Recovery (GL) (Local) Recovery Anticipated SDL Resource Unit Recovery Achieved to Future Total Requirement Date (v2.05 Infrastructure (GL) Factors)* Recovery Paroo 0 0 0 0 Warrego 8 8 0 8 Nebine 1 1 0 1 Moonie 0 2 0 2 Condamine-Balonne 100 58 7.2 65 Queensland Border Rivers 8 15 0 15 Intersecting Streams 0 8 0 8 Gwydir 42 48 0 48 NSW Border Rivers 7 3 0 3 Namoi 10 13 0 13 Macquarie-Castlereagh 65 83 0 83 Barwon-Darling 6 28 3.5 31 Total 247 267 10.7 278 *The to-date recovery numbers were sourced in December 2015, and are based on current planning assumption (i.e. Cap factors version 2.05) which may change in the future. Also, these numbers were rounded to the nearest integer to reflect the accuracy of representing water recovery in the models

Of the nine whole-of-north scenarios, five were based on these principles. But it was recognised that the assumptions regarding the distribution of future water recovery could affect the outcomes achieved by the Basin Plan. For this reason, a subset of specialised scenarios was completed in which the ‘existing + default’ assumption was varied to test the sensitivity of Basin Plan flows to the location and type of recovered water. These specialised scenarios are:  Scenario C — 350 GL — a targeted recovery scenario in which the Condamine– Balonne was not required to contribute to the shared recovery component  Scenario G — 320 GL — the shared recovery was distributed using a near-pure default calculation as described in the Basin Plan (i.e. existing recovery was ignored)  Scenario I — 345 GL — a refined scenario in which the distribution was based on the environmental, social and economic outcomes from Scenarios A – H  Scenario J — 321 GL — a refined scenario in which the distribution was based on the environmental, social and economic outcomes from Scenarios A – I

The stand-alone Condamine–Balonne scenarios (most of which were completed in the Phase I) were testing water recovery parameters including total recovery volume. 5.1.2. Representing Water Recovery in the Models Different approaches were adopted for the unregulated & regulated catchments to represent water recovery in the models. These approaches recognise that location of recovery is likely

Page 46

Hydrologic Modelling for the Northern Basin Review to be an important factor governing flow outcomes in unregulated regions. Note that none of the scenarios included water recovery from the Paroo catchment4. As an over-arching principle, the Basin Plan modelling process ensured SDL compliance was maintained in all catchments. In some cases this required model adjustments to mitigate growth in use (for example, when modelling the recovery of water upstream of Beardmore Dam at St. George on the Balonne River). In practice, state water authorities may follow a different approach to ensure SDL compliance. If so, the year-to-year flow and diversion patterns may be different, but the overall change in flows outcomes is expected to be similar. Gwydir, Namoi, and Macquarie Most of the consumptive use in the Gwydir, Namoi & Macquarie catchments is associated with regulated entitlements. For these regions, water was recovered in the models from regulated entitlements only (the access conditions and volume of unregulated entitlements were unchanged in each scenario). Furthermore, only general security entitlements were subject to the recovery process, hence the volume of high security entitlement was also unchanged in each scenarios. To ensure that the resource assessment included the characteristics of recovered water, a sleeper irrigator containing the recovered entitlement volume was included in the model. This process was modified slightly for the Namoi catchment. Firstly, only users downstream of Keepit Dam were included in the modelled recovery process (i.e. users in the Peel catchment and in the Namoi catchment upstream of Keepit Dam were excluded from water recovery). This modification was made based on feedback from the Commonwealth Government regarding the recovery strategy adopted to date in this catchment. And as a second modification, Scenario I (1103; 345 GL refined) included 4 GL of floodplain harvester recovery between Bugilbone and Walgett to test the effectiveness of these flows to assist with outcomes along the Barwon–Darling. All other scenarios included the recovery of only general security entitlements in the Namoi catchment. The MDBA had incomplete information regarding the specific entitlement holders that had been recovered to date in these catchments. Water recovery was therefore disseminated throughout the catchment, such that the entitlement volume of all general security entitlement holders was reduced by an equal proportion (where the proportion was chosen to meet the desired SDL). This is a modelling assumption that was made for these three catchments, however it is unlikely to affect the modelling results due to the regulated nature of these systems. These catchments contain large upstream storages providing a significant level of flow regulation, from which environmental flows were requested against an environmental account (see Section 5.2). If individual entitlement holders had been recovered in the model (as opposed to the adopted disseminated approach), the environmental demand series would have been the same, and it is expected that the resulting flows through the system would have been near-equivalent.

4 Watercourse diversions from the Paroo system are small (long-term average of 0.2 GL/y), hence there is little scope (and little environmental need) for water recovery from this catchment. Furthermore, the Paroo system only contributes water downstream during extreme flood events, hence water recovered from this catchment would provide only negligible benefit for Barwon–Darling flows. Page 47

Hydrologic Modelling for the Northern Basin Review

Warrego, Condamine–Balonne, and Moonie In unregulated regions (Warrego, Condamine–Balonne, and Moonie), specific entitlement holders were removed from the consumptive pool to achieve the desired reduction in diversions. The unallocated water gifted to the Commonwealth in the Warrego and Moonie catchments was represented by the deactivation of this take in the model. In the Condamine–Balonne, existing recovery was represented by removing those specific entitlements from the consumptive use pool (i.e. existing recovery was represented by both location and entitlement type). Remaining recovery in this system generally followed a targeted approach suitable for each scenario, described further in Section 6. In some cases, the model representation of access rights would result in the remaining users taking a slightly greater volume of water from the river — their utilisation increased in response to increased flows in the river. In these cases, additional entitlements were deactivated to ensure that the target reduction was achieved. In the Lower Balonne, to avoid increased utilisation in another diversion window for the same user, a multi-year volumetric limit was included. This modelling approach ensured SDL compliance, however it is anticipated that a different approach (i.e. not requiring additional recovery) would be followed in practice. This is a key consideration moving forward to ensure compliance with the SDL is observed in practice. Border Rivers The Border Rivers was identified to be a mixed regulated/unregulated catchment. Under baseline conditions, approximately 90% of Queensland Border Rivers take is unregulated, whereas the split in the NSW Border Rivers is approximately 50/50. The representation of Basin Plan water recovery in the models followed these ratios (i.e. a pro rata approach), noting that the Macintyre Brook system was excluded from the water recovery program. That is, the recovery program was divided into four distinct groups: Queensland regulated and unregulated; NSW regulated and unregulated. For both the regulated and unregulated components, the desired reductions were achieved through the deactivation of randomly selected general security entitlement holders. The supplementary cap for NSW was also proportionately reduced to maintain SDL compliance. To ensure that the resource assessment for the regulated component included the characteristics of recovered water, two sleeper irrigators (one each from NSW and Queensland) were allocated regulated recovery volumes from the States respectively. Barwon–Darling Similar to the other unregulated Northern catchments, the representation of water recovery in the Barwon–Darling was achieved through the removal of specific entitlement holders from the consumptive pool. As an additional step, the method was refined to ensure that the proportion of water recovered from each of the three main reaches (Mungindi to Walgett, Walgett to Bourke, and Bourke to Wilcannia) followed the baseline split (i.e. a pro rata recovery at the reach-scale). This was a modification to the pumping threshold-based recovery mechanism adopted during Basin Plan development modelling (MDBA 2012a), and was made based on feedback from NSW modelling officials, and to ensure a more realistic representation of the

Page 48

Hydrologic Modelling for the Northern Basin Review water recovery process. Furthermore, adopting this approach ensured a consistent water recovery methodology was followed for all Northern Basin unregulated catchments. Intersecting Streams The water recovery program achieved to date includes 8 GL of recovery from the Intersecting Streams SDL resource unit, the result of the purchase of Toorale Station by the Federal and New South Wales Governments in 2008. The geographic location of Toorale allowed irrigation water to be extracted from both the Warrego and Barwon-Darling catchments. The Intersecting Streams region is not explicitly included in the MDBA modelling framework, hence this component of recovered water was divided between the Warrego and Barwon–Darling models. As such, the 6 GL/y of water extraction entitlements that are represented in the Barwon–Darling model, but part of the Intersecting Streams SDL resource unit were deactivated. These entitlements were associated with approximately 4 GL/y of consumptive use, but were not counted towards the Barwon–Darling BDL. Furthermore, the inflows from the Warrego model were scaled to provide a 2 GL/y long-term increase. 5.2. Representing Environmental Water Use Modelling a fully implemented Basin Plan also required assumptions about future environmental watering practices. For unregulated regions the approach was to assume that long term average flow would increase through compliance with the SDL. The process was more complex for regulated rivers. In these catchments, water recovery makes up a share of the available water resource in storage, for which the environmental water holder must choose a desired pattern of releases. The process of making environmental water releases includes a variety of practical considerations such as allocation announcements, existing flows in the river, channel sharing arrangements, system constraints, and so on. It is not clear how this process will develop as the Basin Plan continues to be implemented over coming years. The ESLT method (MDBA 2011) adopted for the Basin Plan used environmental science as a foundation for the assumed modelled watering strategy. Under this approach, the pattern of environmental water requested from upstream storages in the model was based on the site-specific flow indicators, where these indicators represented broader environmental water requirements of river valleys or reaches. The pattern of environmental water requested from upstream storages in the model were defined using these SFIs. This method was used during the modelling completed for Basin Plan development (MDBA 2012a), and has been continued in the Northern Basin Review, albeit using updated environmental science (MDBA 2016c). This assumption was made with a view towards the long-term development of environmental watering strategies. Environmental watering has been occurring in the Northern Basin over the past few years, but it is a relatively young process. The Commonwealth Environmental Water Office (CEWO) is still learning the most effective methods to deliver environmental flows, and the Basin Plan modelling methodology required a prediction of where these learnings will lead in decades to come. The modelling therefore assumed that:  this learning process will continue;  increased interactions between environmental water holders and river operators will improve the ability to supplement existing flows with environmental releases, and; Page 49

Hydrologic Modelling for the Northern Basin Review

 future advances in predictive capacity will increase the efficiency and capacity to coordinate environmental water delivery. ‘Environmental demand time series’ provided the mechanism by which environmental watering was included in the model. These demand series requested a specific pattern of flows to be delivered at a certain site over the 114-year modelling time period, where this pattern was driven by the SFIs. In the vast majority of cases, the demand series requested storage releases to supplement existing events in the system — that is, demand series did not seek to build entire events through releases. This is consistent with the watering approaches that have occurred historically. Also, as the demand series were based on the SFIs, each requested flow event can be associated with a specific set of environmental outcomes. Demand series were included in the Border Rivers, Gwydir, Namoi, and Macquarie catchments, as these rivers contain a significant capacity to regulate flow through public dams. Also, demand series were included at multiple locations in each of these catchments, each associated with different aspects of water use. For instance, environmental watering in the Namoi model included demand series for baseflows, a demand series to request environmental flows for riparian outcomes in the Lower Namoi, and another demand series to request flow events to enhance environmental outcomes downstream in the Barwon– Darling. A more detailed description of the demand series approach for Basin Plan modelling is given by MDBA (2012a). A few of the model scenarios also included a demand series in the Condamine–Balonne. In these scenarios, regulated water was recovered from the St George region, and an associated demand series was placed at Jack Taylor Weir to replenish low flows through the Lower Balonne. This was not included as standard practice, but instead as a proof-of- concept to test the capacity for releases from Beardmore Dam (through Jack Taylor Weir) to enhance low flow outcomes. The specific scenarios including this demand are described in Section 6. The methods and approaches used to produce the demand series were found to be consistent with the requirements of the Basin Plan, and are technically sound and fit-for- purpose (Bewsher 2016). 5.2.1. Environmental Water Accounting in the Models The baseline models include a simulation of the water accounting and allocation mechanisms that occur in practice, whereby water orders from storage and extraction against an entitlement are subject to the availability of the resource. Of key importance to Basin Plan modelling is the assumption that the environmental water holder is (and will be) subject to the same water management arrangements as all other users in the system. That is, the entitlements recovered for environmental use provide the same access rights, and are subject to the same flow constraints and allocation announcements as all other entitlements. However, the Northern Basin models do not include an environmental account for water recovered under the Basin Plan. It was identified early in the Basin Plan modelling process that environmental water use without an accounting filter could significantly depart from current sharing arrangements and unintentionally impact the access rights of other users in the model. Including Basin Plan environmental accounts would therefore be desirable,

Page 50

Hydrologic Modelling for the Northern Basin Review however it would require a significant dedication of resources, both from MDBA and State modelling representatives, and was therefore not completed for the NBR. The MDBA therefore included two components in the modelling process to ensure that the access rights of other users in the model were maintained over the long-term. The first was a model-external accounting mechanism (the EEST; see below) to enable the environmental watering strategy to be represented in the modelling framework, and the second was the iterative modelling process. The first of these steps provided an initial check of environmental water use based on annual accounting estimates; the second step refined this estimate using the more accurate daily flow routing calculations in the model. 5.2.2. Demand Series for Local Outcomes Each demand series was generated external to the modelling framework. The SFIs underlying each demand series are described further in MDBA (2016c), and are summarised in Table 7.

Table 7: Site-specific flow indicators underlying the demand series for local needs

Catchment Site-Specific Flow Indicator (Demand Location) 4,000 ML/d for 5 days (Oct – Dec) Border Rivers 4,000 ML/d for 5 days (Oct – Mar) (Mungindi) 2 events of 4,000 ML/d for 11 days (all year)

Gwydir 4,500 ML/d for 92 days (Nov – Jan) (Mallowa offtake) 5,400 ML/d for 120 days (Feb – Aug & Mar –Sep) 150 ML/d for 45 days (Oct – Jan) 1,000 ML/d for 2 days (Oct – Jan) 45 GL within 60 days (Oct – Mar) Gwydir 60 GL within 60 days (Oct – Mar) (Yarraman Bridge) 80 GL within 60 days (Oct – Mar) 150 GL within 60 days (Oct – Mar) 250 GL within 60 days (Oct – Mar) 500 ML/d for 75 days total (all year) Namoi 1,800 ML/d for 60 days total (all year) (Bugilbone) 4,000 ML/d for 45 days total (all year) 100 GL within 5 months (Jun – Apr)

Macquarie 250 GL within 5 months (Jun – Apr) (Marebone break) 400 GL within 7 months (Jun – Apr) 700 GL within 8 months (Jun – May)

To ensure that the water access conditions for users were maintained over the long-term in Basin Plan scenarios, the MDBA developed a model-external accounting mechanism for environmental water use — the Environmental Event Selection Tool (EEST; MDBA 2012a). Page 51

Hydrologic Modelling for the Northern Basin Review

The overall purpose of the EEST is to allow the user to build a 114-year environmental watering pattern to be included in the modelling framework. The tool included two elements. Firstly, it contained a yearly volume of available environmental water, providing an annual first-pass accounting mechanism (which was then refined through the iterative process described below). Secondly, the tool included a number of possible environmental flow events that was available for re-instatement in each year. Under this process, an ‘available event’ refers to a flow event that occurred in the without development conditions but was no longer present under baseline conditions. Combined, these two elements filtered the available environmental flow events to an annual estimate of the environmental account volume. The State hydrological models then determined which of these events it was possible to deliver in practice based on more accurate estimates of available environmental water and system losses. Border Rivers, Namoi, and Macquarie Demand Series An EEST was built for the Border Rivers, Namoi, and Macquarie catchments (the Gwydir catchment is described later) by identifying those events available for re-instatement, completed through the following steps. 1. The baseline flow time series was analysed for each indicator site and each water year (from 1895 to 2009) to identify all existing flow events that satisfied the SFI criteria listed in Table 7 2. The without development time series was similarly analysed to identify all successful SFI events 3. A comparison of the outcomes from steps 1 and 2 was used to identify the environmental flow events that have been lost from the without development time series due to river regulation and extraction. This identified eligible events that could be reinstated in the model to meet the desired frequency of environmental watering 4. The volume of water required to reinstate each individual lost event in the without development timeseries was then calculated (see the example in Figure 15) 5. Events that successfully occur in both the baseline and without development model scenario are considered to be ‘achieved by existing flows’ and are not available for selection 6. The volume of water expected to be available to the environmental water holder in each year was calculated for the given Basin Plan model scenario (representing a particular SDL option). This annual environmental account followed the same annual pattern as baseline diversions, but scaled down for the appropriate SDL based on the proportion of regulated entitlements held by the environmental water holder. These steps provided a complete EEST for each of the four regulated catchments. During Basin Plan modelling (MDBA 2012a) a set of principles were developed governing the use of the EEST to ensure that the chosen pattern of events had an environmentally sound basis. These principles (described below) were designed to reflect the environmental water manager and river operator decision-making processes that occur in practice.  Water use must remain within the EEST environmental account (with a 20% allowance in some years to reflect carryover of environmental allocation; see below).  The frequency of selected events will be driven by the target frequencies for each SFI (see MDBA 2016c for a description of these frequencies).

Page 52

Hydrologic Modelling for the Northern Basin Review

 The user will aim to minimise the length of dry periods (i.e. periods between watering events). The primary factor limiting the selection of events was the annual account volume. That is, as a general rule, all events with a volume lying within the account volume were activated. In a small number of years (i.e. once or twice over the 114-year modelling period), the user was required to choose between multiple events — the frequency and dry spell criteria were used to prioritise these events5. But overall the minimal degrees of freedom meant that the spectrum of possible event sequences for each Northern Basin scenario was very narrow. The 20% allowance in the water account is a modelling representation of the environmental water holder having access to the same carry-over provisions as other entitlement holders in the system. This allowance is a long-term estimate based on historical allocations to general security entitlement holders. The EEST calculations allow the user to exceed the account by 20%, but with the associated requirement that the rolling average annual water use remains within the account (i.e. any carry-over use is re-balanced by reduced access in surrounding years). This allowance was accessed in only small minority of years and is tempered by the hydrological models ability to determine what could actually be delivered. The first scenario completed for NBR was Scenario B (390 GL; a fully implemented Basin Plan under its current settings). Events were selected for this scenario using the volume, frequency and dry spell principles detailed above. For all subsequent scenarios, the Scenario B event sequences were modified based on the updated recovery volumes, and the event sequences were updated based on the new annual account volume. A summary of the number of events included in the demand series for each scenario over the 114 years is provided in Table 9. As an example of this process, Scenario A (415 GL) represented a larger volume of water recovery, resulting in the Macquarie environmental water account increasing from 83 to 88 GL (Table 9). In this case, the Macquarie EEST was modified to include the increased annual environmental account volume, and all additional events lying within the increased volume were activated. Conversely, the Macquarie account volume in Scenario J was reduced (to 55 GL; Table 9), resulting in the deactivation of those events that no longer satisfied the accounting rules. The only exception to this approach is provided by Scenario G for the Macquarie model. In this scenario, the number of requested events was greater than Scenario B, despite a decrease in recovery volume. This special case is due to additional events selected as part of the iterative process to ensure each model run correctly replicated the desired SDL. See section 4.2.4 for a detailed explanation of this process. This was a pragmatic decision to modify the environmental demand series rather than other model parameters such as irrigable areas as environmental demands are the most uncertain variable in the modelling process.

5 For context, the selection of events in the Southern Basin EEST follows the same process and principles, but the frequency and dry spell criteria have a greater influence on the event selection outcome. This is a result of the larger number of available environmental events, and the larger volume of regulated environmental water in the Southern Basin. Page 53

Hydrologic Modelling for the Northern Basin Review

12,000 Additional Flow Requested from Storage

10,000 Flow Prior to Basin Plan (Baseline Conditions)

Without Development Conditions 8,000

6,000

4,000 Flow at Bugilbone Bugilbone Flow(ML/day) at

2,000

0 Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Figure 15: Example environmental demand series, showing a requested flow event in the Lower Namoi (measured at Bugilbone) over a 12-month period. The blue region shows the additional environmental flow requested for delivery to supplement existing flow events. Figure 15 displays an example demand series in the Lower Namoi (measured at Bugilbone) for a 12-month segment of the model series. The demand series is marked in blue and baseline flows are marked in red. This example is emblematic of most requested events, in that it has requested storage releases to supplement and existing flow event and restore a specific part of the without development flow. The indicators for the Border Rivers and Namoi require the flow to exceed a given threshold for a number of days, hence they are flow-based. In contrast, the SFIs for the Macquarie Marshes are volumetric, hence a successful event will achieve the desired volume within the set period, regardless of flow. The demand for the Macquarie Marshes was limited to a flow of 4,000 ML/d, as this was the constraint operational during the development of the Basin Plan, and has been carried through to the Northern Basin Review. The MDBA are aware that large flows in recent years has led to a developing breakout near Mumblebone (upstream of Marebone), and that this has also reduced the operational constraint. The most recent information indicated a new constraint of 3,200 ML/d. However, as this situation is still developing, the updated constraint information has not been included in the model (both for irrigation and downstream demands). Initial testing indicates that, as the requirements for the Macquarie Marshes are volumetric (i.e. not flow-based), this constraint would have only minor effects on the delivery of the desired events. Gwydir Demand Series Demand series were developed for two Gwydir sites. The Mallowa Creek demand series followed the method described above, and were unchanged from those used as part of Basin Plan development (MDBA 2012a). These demands aimed to restore volumes of 4.5 GL (over three months) at the Mallowa Creek regulator. Demand series for the second site (Gwydir wetlands) followed an alternative method. The modelling approach adopted prior to 2012 made use of the existing environmental contingency allowance (ECA) in the baseline model, but modified to meet Basin Plan objectives (MDBA 2012a). This method contained some weaknesses, in that recent adaptations in ECA management are not represented in the baseline model, and the Page 54

Hydrologic Modelling for the Northern Basin Review modifications introduced potential flooding impacts downstream of Yarraman Bridge. Hence this approach was not adopted for the Northern Basin Review. Instead, an updated approach was developed that recognised the distinct characteristics of the catchment, its operational constraints, and the environmental asset. A dominant feature of the catchment is Copeton Dam (completed in 1976), a substantial regulating structure located towards the headwaters of the Gwydir River. It is a relatively large storage given the upstream catchment area, capturing and regulating all inflowing water (to date, the storage has not spilled). About half of the inflows to the total Gwydir catchment are captured by the storage — downstream of the dam, the Gwydir River continues to receive unregulated inflow events from several creeks. The presence of Copeton Dam has removed or reduced many of the large inflow events to the main river stem. Furthermore, extraction from the river system has further reduced the volume of water passing through the catchment to the Gwydir wetlands. The wetlands continue to receive flows, but they have been much reduced by development. The SFIs developed to monitor inundation of the Gwydir wetlands (MDBA 2012b) were specified in terms of total volume over a period of 60 days. The demand series approach adopted for the Gwydir wetlands aimed to supplement the remaining unregulated flow events with releases from Copeton Dam to achieve the volumes specified by the SFIs, where these releases were constrained by the relatively low capacity of the river channel in the Lower Gwydir (private landholdings adjacent to the Gingham Watercourse can be inundated at relatively low flows). The flow limit was set to 300 ML/d at Yarraman Bridge, consistent with existing operational practices. This is a different approach to the other Northern catchments. In the Border Rivers, Namoi and Macquarie, the modelling approach aimed to restore ecologically significant parts of the without development flow regime. In the Gwydir, the modelling approach recognised that existing operational constraints in the Lower Gwydir prohibit the restoration of natural flow regime elements. Instead, environmental outcomes were targeted by providing the target volume of water at low flows (300 ML/d) over a significantly longer period than would have occurred under the natural flow regime. The five inundation SFIs for the Gwydir wetlands are summarised in Table 7. A successful event will achieve the desired volume within a 60-day window. Similar to other catchments, the Gwydir demand series were built to request the delivery of events with an SFI shape, but subject to river operator and environmental water manager decision making processes that could occur in practice. In the Gwydir, this was achieved by using natural inflows to trigger the release of environmental water. Following a typical large rainfall event, the unregulated creeks downstream of Copeton Dam generally contribute inflows in the form of a peaked event with the majority of flow falling within seven days. Seven-day events at Yarraman Bridge were therefore used to trigger events in the Gwydir wetlands demand series. This is a modelling representation of an approach that could plausibly occur in practice, whereby environmental water managers and river operators monitor predicted inflows from downstream creeks and make environmental releases from Copeton Dam to supplement the ‘tail’ of these events. A threshold for the trigger event was determined for each SFI, listed in Table 8. The trigger volumes were chosen based on an examination of the baseline time series at Yarraman

Page 55

Hydrologic Modelling for the Northern Basin Review

Bridge. The maximum trigger volume for each year in the baseline scenario was found and correlated with the successful achievement of SFIs. This gave an indication of the probability that a particular seven-day event would be followed by enough flow to satisfy the SFI criteria.

Table 8: Seven-day event volumes used to trigger environmental releases for each Gwydir wetlands SFI

VOLUME AT YARRAMAN BRIDGE (GL) Gwydir Wetlands 7-day Trigger SFI Volume 45 9 60 10 80 14 150 50 250 70

Under baseline conditions, the trigger volumes listed in Table 8 led to a successful SFI event on 20 to 40% of occasions (depending on the SFI). The purpose of the demand series was to increase this rate of achievement by adding supplementary environmental releases to follow each seven-day trigger event. The chosen trigger volumes listed in Table 8 represent a balance. Increasing the threshold volume would have provided a greater confidence that supplementary environmental flows would achieve the desired volume, but it would have also decreased the frequency of opportunities. The operational constraint in the Lower Gwydir (300 ML/d at Yarraman Bridge) limited the rate of environmental releases requested by the demand series. Furthermore, as each SFI is associated with a period of 60 days, the seven-day trigger events were supplemented with environmental releases lasting 53 days (but subject to the October–March seasonality required by each SFI). An example of a desired event is shown in Figure 16. As a final step, to ensure consistency with the overall Basin Plan modelling approach, the environmental releases in the Gwydir were subject to a model-external environmental account — the EEST process, combining the two accounting elements (an annual environmental account and an estimated cost for each event). Consistent with the other EESTs:  the annual Gwydir account followed the same pattern as baseline diversions, but scaled down to the appropriate recovery volume; and,  the ‘cost’ to deliver each proposed environmental event was measured to be the difference between the demand series and baseline flows (i.e. the difference between the orange a blue lines in Figure 16). If the cost of an event exceeded the account volume, the accounting mechanism limited the duration of the environmental releases. That is, environmental releases would be requested, but not for the entire 53 days. This is likely to affect the success rate of the requested events, but it was included to reflect the accounting mechanisms for water use that occur in practice. Consistent with the EESTs for the other catchments, this accounting mechanism included an additional 20% allowance to represent carry-over of environmental water to the next year.

Page 56

Hydrologic Modelling for the Northern Basin Review

As a result, the selection of events for the Gwydir did not change between scenarios. The events were chosen through the standardised trigger-based process described above, hence the number of events listed in Table 9 for Gwydir inundation events is constant. The difference in demands between scenarios was the volume of environmental releases that could be dedicated to each event, determined by the recovery volume.

Figure 16: Example demand series event for the Gwydir Wetlands, measured at Yarraman Bridge. The orange line traces the requested flows to supplement the initial trigger event. The Gwydir demand series approach is a standardised method that includes only a small amount of user judgement. The single subjective element is the choice of threshold for the trigger events. The only effect of altering this threshold would be to change the number of events included in the demand series — the pattern of these events over the 114-year modelling period would still be determined purely by the analytical steps described above. Furthermore, the success of these requested events would be determined by the model, which attempts the deliver the requested flows subject to its representation of operational factors such as storage release capacity, channel sharing requirements, conveyance losses, and constraints. Baseflow Demand Series Each of the four regulated models also included baseflow demand series at key locations (see Appendix B for the specific sites). These series requested low flow releases from storage over the 114-year period, and are associated with the maintenance of aquatic habitats for fish, plants and invertebrates. These series comprised only a small proportion of the total environmental water use in the four regulated catchments. The method used to build these demand series is described in the Basin Plan hydrologic modelling report (MDBA 2012a), and was not changed for the NBR. Baseflow results are summarised in Appendix B.

Page 57

Hydrologic Modelling for the Northern Basin Review

Table 9: Number of events included in the 114-year demand series for local environmental requirements in each scenario

Scenarios Model Scenario Scenario D Scenario E Scenario G Scenario J Scenario I Scenario C Scenario A Catchment Site B & H Whole-of-north Recovery (GL) 278 320 320 PR 321 345 350 390 415 Recovery (Regulated Component) (GL) 18 (3.4) 28 (6.3) 34 (9.6) 28 (6.3) 42 (7.9) 41 (11.4) 39 (11.2) 43 (12.7)

Border 4,000 ML/d for 5 days (Oct – Dec) 0 0 4 0 4 4 4 4 Lower Number of Rivers 4,000 ML/d for 5 days (Oct – Mar) 0 0 1 0 1 1 1 3 Macintyre Requested Events 2 events of 4,000 ML/d for 11 days (all year) 2 3 5 3 5 5 5 6 Recovery (GL) 48 48 51 47 47 56 56 59 150 ML/d for 45 days (Oct – Jan) 0 (achieved through other watering actions) 1,000 ML/d for 2 days (Oct – Jan) 0 (cannot be actively managed within current constraints) 45 GL within 60 days (Oct – Mar) 13 Gwydir Number of 60 GL within 60 days (Oct – Mar) 10 Gwydir Wetlands Requested Events 80 GL within 60 days (Oct – Mar) 15 150 GL within 60 days (Oct – Mar) 11 250 GL within 60 days (Oct – Mar) 6

Mallowa Number of 4,500 ML/d for 92 days (Nov – Jan) 57 Creek Requested Events 5,400 ML/d for 120 days (Feb – Aug & Mar –Sep) 0 (WSP replenishment flow included in the model) Recovery (GL) 13 20 20 20 24 24 24 28 500 ML/d for 75 days total (all year) 6 7 7 7 7 7 7 7 Namoi Lower Number of 1,800 ML/d for 60 days total (all year) 0 0 0 0 0 0 0 0 Namoi Requested Events 4,000 ML/d for 45 days total (all year) 4 7 7 7 7 7 7 7 Recovery (GL) 83 83 77 55 74 83 83 88 100 GL within 5 months (Jun – Apr) 7 7 7 2 6 7 7 8

Macquarie Macquarie Number of 250 GL within 5 months (Jun – Apr) 13 13 16 9 11 13 15 15 Marshes Requested Events 400 GL within 7 months (Jun – Apr) 6 6 7 4 7 6 8 10 700 GL within 8 months (Jun – May) 1 1 1 0 1 1 1 1

Page 58

Hydrologic Modelling for the Northern Basin Review

5.2.3. Demand Series for Downstream Outcomes The Basin Plan includes a volume of water recovery to achieve environmental outcomes in the Barwon–Darling. This volume is shared throughout the Northern Basin — it recognises that inflows from local rainfall in the Barwon–Darling are relatively small, hence flows through this system are almost entirely reliant on the combined inflows from upstream tributaries. For the purpose of Basin Plan modelling, it was assumed that future environmental watering strategies for each catchment will target local outcomes as a first priority (described in the previous section), but it was also assumed that water remaining in the environmental allocation would be dedicated to enhancing downstream outcomes. This downstream component was represented through a set of demand series located towards the lower end of each catchment with a significant flow regulating capacity (i.e. the Border Rivers, Gwydir, Namoi & Macquarie; demand locations are listed in Table 10). The capacity to which environmental water holders will be able to deliver downstream water in the future is unclear. As technology and flow predictive capacity improves, future downstream watering strategies may include a cooperative element in which the regulated catchments coordinate their downstream releases to enhance Barwon–Darling outcomes. The efficient use of environmental water for whole-of-system outcomes is an overarching principle of the Basin Plan, but in the northern Basin it represents a change (to some degree) to existing operational capacity. Current operational practices have not been designed to deliver regulated water through a catchment and downstream to the Barwon– Darling. And from a practical perspective, flows through the Barwon–Darling usually require contributions from multiple catchments, but the unique geomorphology of each catchment, in conjunction with the highly variable nature of the climate and river flows, introduces a relatively large degree of uncertainty to the forecasted travel times and losses of individual flow events. For this reason, the MDBA have modelled two possible strategies.  Strategy 1 (whole-of-north coordinated) represented a highly-managed system in which a two or three catchments work proactively to coordinate flows into the Barwon–Darling. This would require improvements in forecasting tools and experience, and substantial changes to existing operational practices.  Strategy 2 (catchment-scale), in which environmental releases were made in each catchment individually to maintain low flow targets in the Barwon–Darling. The adopted low flow targets were drawn from the Interim North-West Unregulated Flow Management Plan (Barwon–Darling Water Sharing Plan). This catchment-scale management strategy more in-line with current operating practices. Actual watering strategies to manage the water recovered across the northern Basin are still undergoing development, and it is not yet clear the level to which cross-catchment coordination will be implemented in practice. The two modelled strategies together encompass a broad range across the ‘forecasting and coordination spectrum’. Both strategies are considered to be achievable in practice, but Strategy 1 would require a larger commitment of investment towards flow forecasting and coordination capacity over coming years. The maximum target flows during downstream delivery periods is listed for both strategies in Table 10. Note that these were the maximum requested flows in each demand series. For Page 59

Hydrologic Modelling for the Northern Basin Review most events, the actual requested flow was less. Furthermore, the demand series were inputs to the modelling framework. The model then attempted to deliver the requested daily flows, but would do so subject to more accurate representations of losses, flow travel times and operational constraints. The maximum requested flows for Strategy 1 were chosen based on the bankfull level at each location, and incorporated the connectivity work described in section 3. Specifically, the maximum requested flows recognised that the Namoi and Border Rivers have the greatest capacity to provide higher peaked flows due to their relatively high level of downstream connectivity. The flows chosen for Strategy 2 were based on the baseflow targets listed in the Interim Unregulated Flow Management Plan for the North West, described further below.

Table 10: End of system locations and maximum target flows for the regulated catchments in both the coordinated and not coordinated model scenarios

Upper Limit for EOS Flows (ML/d) Catchment Strategy 1 Strategy 2

Border Rivers (Mungindi) 4,700 850

Gwydir (Collarenebri) 300 300

Namoi (Goangra) 4,000 700

Macquarie (Marebone Break) 4,000 4,000

The method for determining the Environmentally Sustainable Level of Take (ESLT) is based on the five “ecologically significant components of the flow regime” (see MDBA 2011 for more information). Ranked in order of lowest to highest flow event, these components are cease-to-flow events, base/low flow events, freshes, bankfull flows, and overbank flows. As a general rule, the highest flow events also comprise the largest volumetric contribution to the ESLT. Due to their different underlying principles (described in more detail below) the downstream watering strategies targeted different ESLT components of the flow regime. Strategy 1 targeted freshes, while Strategy 2 was aimed towards cease-to-flow and baseflow events. That is, Strategy 1 used key SFIs to determine the pattern of releases from storage (Strategy 2 was not SFI-based), and targeted flows with the greatest sensitivity to the ESLT. For this reason, Strategy 1 is the most consistent with the overall ESLT method. The outcomes of both options were presented to the Authority to inform the triple-bottom line assessment of SDL options in the Northern Basin, but Strategy 1 was adopted as the standard approach for most model scenarios presented in this document. Strategy 1 — Whole of North Coordinated Releases Downstream demand series for Strategy 1 were developed using a Barwon–Darling EEST. This tool followed the same principles as the EESTs developed for local requirements in the Northern tributaries, but refined to reflect to unique nature of the Barwon–Darling system.

Page 60

Hydrologic Modelling for the Northern Basin Review

Coordinated releases were activated to target flows between 6,000 and 10,000 ML/d at Bourke. These flows correlate to the SFIs coloured green in Table 11. The remaining SFIs are associated with large flow events that were deemed to be beyond the influence of regulated flows.

Table 11: SFIs in the Barwon–Darling — indicators targeted through the Strategy 1 (coordinated releases) are coloured green

Flow Location SFI 6,000 ML/d for 14 days (all year) 10,000 ML/d for 14 days (August - May) Two events of 10,000 ML/d for 20 days (August - May) Bourke 30,000 ML/d for 24 days (all year) 45,000 ML/d for 22 days (allyear) 65,000 ML/d for 24 days (all year) 6,000 ML/d for 20 days (August - May) Louth 21,000 ML/d for 20 days (August - May) Two events of 6,000 ML/d for 7 days (all year) Wilcannia 20,000 ML/d for 7 days (all year) Annual flow volume of 2,350 GL when flows are >30,000 ML/d

Similar to the other tributary EESTs, the Barwon–Darling version of the tool sought to re- instate those events that had been lost from the without development time series due to river regulation and extraction. The Barwon–Darling EEST also included an estimate of the increased flow at Bourke that would result from Basin Plan implementation in the unregulated catchments, an acknowledgement that a significant proportion of Barwon– Darling flow is sourced from unregulated rivers. This approach represents a management strategy in which storage releases would be triggered by forecasted flows at Bourke. To achieve this type of strategy in practice, river operators would require the ability to aggregate current flows in the tributaries to forecast flow at Bourke (with some accuracy) over the next few weeks. This level of predictive capacity does not yet exist for this system but is expected to evolve over time as operators adapt to delivering environmental water in the future. The Barwon–Darling EEST was subject to significant upgrades from the version used during Basin Plan development. The overall outcome of these upgrades was to:  allow the requested end of system flow to vary from event to event based on existing flow conditions in each catchment and the water available in the environmental account each year, and;  standardise the selection of target downtream events. The following steps were followed to construct the downstream demands: 1. An interim model scenario was completed in which the Basin Plan was modelled in the unregulated catchments (i.e. Paroo, Warrego, Condamine–Balonne, Moonie, and Barwon–Darling), but the remaining catchments remained at baseline settings.

Page 61

Hydrologic Modelling for the Northern Basin Review

2. The interim scenario was analysed to identify all existing flow events that achieved the target flow events. 3. A comparison of this scenario to the without development scenario revealed those events at Bourke that had been removed due to river regulation and extraction, and had not been restored by implementing the Basin Plan in the unregulated catchments. These events were identified to be opportunities for supplementing regulated releases. 4. An estimate was made of the increase in end of system flows that could be achieved in the regulated catchments during each of the opportunities (typically a two week period). This determination was calculated to be the minimum of the following: a. without development flows during the desired period; b. an estimate of bankfull at the demand location (Table 10); c. the flow that could be achieved given the environmental account volume in that year. 5. For each of the identified opportunities, the estimated contributions from the regulated catchments were added to the interim scenario flows. This provided a first cut determination of whether regulated releases could add enough flow to achieve the SFI. 6. Those events for which the regulated releases were estimated to supplement existing flows to at least 80% of the target volume were selected for the demand series (this 80% cutoff value was chosen based on the 10% allowance in the relevant flow criteria; satisfying 90% of the duration and 90% of the threshold — when combined, around 80% of the volume — are counted as a successful event in Basin Plan modelling; MDBA 2012a). 7. The selected events were translated into daily demand series for each site using the method in step 4. This method is outlined graphically in Figure 17. Similar to the tributary EESTs, the minimal degrees of freedom in the Barwon–Darling EEST meant that the spectrum of possible event sequences was essentially zero — steps 5 and 6 standardised the event selection process.

Page 62

Hydrologic Modelling for the Northern Basin Review

4. Calculate event-by-event contribution from each catchment 1. Interim scenario Bourke flows Annual Water Accounts

Baseline Flow 2 & 3. Find removed Bourke events Without Development Flow

5. Estimate Bankfull achievability of each Estimates Bourke event (see table)

7. Pattern translated 6. Select events to tributary achievability ≥ 0.8 demands

Figure 17: Decision tree determining events to be selected under Strategy 1 (coordinated releases) This method was applied to all four regulated catchments in the Northern Basin, noting that the Gwydir downstream demand included an additional demand component. This was an effect of the operating constraints in the Lower Gwydir, under which the environmental demand series was capped to 300 ML/d (measured at Yarraman Bridge; see section 5.2.2). As a result of this constraint, the regulated contribution of the Gwydir during downstream demand periods was small, hence there was additional capacity within the environmental account to achieve additional outcomes. This capacity was used to deliver supplementary low flows (also capped to 300 ML/d) to enhance downstream connectivity with the Barwon–Darling. These low flows formed the sole downstream contribution from the Gwydir as part of Strategy 2, and are therefore described below. Strategy 2 — Catchment-Scale Releases To test the influence of coordinated environmental watering on flows in the Barwon–Darling, a model scenario was completed in which the coordinated aspect had been removed. Also, given the practical challenges of implementing a coordinated strategy, the catchment-scale releases under Strategy 2 were designed to operate within the bounds of current river operations practice. That is, implementing this strategy would not require the significant development of operator experience, new knowledge and/or technologies. Downstream demand series for Strategy 2 were developed without using a Barwon–Darling EEST. The demand series, although individually constructed for each catchment, all followed the five principles set out below. 1. There was no coordination of environmental demands between catchments. Page 63

Hydrologic Modelling for the Northern Basin Review

2. In-valley environmental demands were the first priority; followed by downstream end of system demands. 3. Downstream flow targets were based on Interim Unregulated Flow Management Plan for the North West, listed in Table 12. 4. Flow targets were also restricted by the operational constraints in each catchment. 5. Demands were limited to daily without development flow (except in the highly modified Gwydir system).

Table 12: Low flow targets listed in the Barwon–Darling water sharing plan under the Interim Unregulated Flow Management Plan for the North West

Location Target (ML/day) Mungindi 850 Collarenebri 760 Walgett 700 Brewarrina 550 Bourke 390 Louth 280 Wilcannia 150

The flows listed in Table 12 are existing low flow targets in the Barwon–Darling water sharing plan. Only this low flow aspect of the Interim unregulated flow management plan was included in the modelling, the higher flow aspects were not adopted. The plan anticipated that these targets could be achieved through an ad hoc embargo on take in the Barwon– Darling — Strategy 2 instead postulates that these flows could be achieved through regulated releases in upstream catchments. These demand principles above were codified into a number of rules for each tributary to generate downstream demand series. Apart from the lack of coordination between catchments, this facet is the most significant difference between the two strategies. The rules-based approach removed the need for a decision tree to govern event selection.

Strategy 2 — Namoi & Border Rivers Implementing the downstream environmental watering strategy in well-connected tributaries (the Namoi and Border Rivers) was relatively straight forward as it simply involved restoring without development end of system flows up to the local Barwon-Darling flow target.

Page 64

Hydrologic Modelling for the Northern Basin Review

Figure 18 Stylised end of system hydrograph of uncoordinated downstream demand series generation in the Border Rivers. The shaded area represents the additional flow requested from storage. The approach adopted for the Namoi and Border Rivers catchments is displayed in Figure 18. The demand series (green line) represents a strategy in which the environmental water holder requests releases to restore without development flows, but only up to the local flows target (850 ML/d in the Border Rivers and 700 ML/d in the Namoi; see Table 10). The additional requested flow (i.e. supplementing baseline flow) is marked by the blue shaded area.

Strategy 2 — Gwydir As the flow constraint in the Gwydir catchment (300 ML/d) is below the local Barwon-Darling flow target (750 ML/d at Collarenebri) the downstream demand series was capped at a flow of 300 ML/d. Also, these flows were delivered during years when there were no Basin Plan demands for the Gwydir Wetlands.

The downstream demand series for the Gwydir requested low flows (300 ML/d at Yarraman Bridge) to pass through the Mehi River, providing additional downstream connectivity. These flows were requested in years without demands for the Gwydir Wetlands, and operated during the August to November period to minimise potential channel sharing constraints with other users in the system.

Strategy 2 — Macquarie-Castlereagh The most significant difficulty in delivering downstream environmental water in the Macquarie- Castlereagh catchment is effectively being able to deliver water through the Macquarie Marshes. If the Marshes are in a relatively dry condition they will absorb the entering flow. As such, the catchment-scale strategy demand series in the Macquarie-Castlereagh was designed to be active only when existing flows were already passing through the Marshes.

Page 65

Hydrologic Modelling for the Northern Basin Review

Figure 19 Schematic diagram for the Macquarie–Castlereagh The Macquarie demand series aimed to supplement existing flows through the Marshes with an additional regulated release, extending the period of connection downstream with the Barwon–Darling. For these releases to be effective, the marshes must be ‘pre-wetted’, such that additional inflows will provide additional outflows (see the connectivity discussion in section 3). As shown in Figure 16, Marebone Break can be considered to be an inlet to the Macquarie Marshes and flows past Carinda an outlet. A comparison of Macquarie River flows at Marebone Break and Carinda was taken as a proxy for the condition of the marshes. That is, if the flow at Carinda exceeded that at Marebone Break, the marshes were considered to be draining, and the demand series was activated. Making environmental releases under these conditions is the most efficient way to pass flows through the marshes without exceeding the channel capacity at Marebone Break. This approach is considered to be conservative as there are alternate flow paths through the marshes that do not pass Carinda, but will be better tracked in updated versions of the Macquarie model that explicitly represent water flow through the marshes. An example of the Macquarie demand series technique is shown in Figure 17 where the shaded area represents the additional flow on top of baseline requested by the demand series. If Macquarie River flow at Carinda flow exceeded Marebone Break flow then the demand series value for that day was set to be the minimum of the channel capacity and without development flows at Marebone Break. This rule set created a demand series in which the flow decreased at the same rate as without development flow for the period in the baseline model run that flows at Carinda exceeded those at Marebone Break. Consistent with the Macquarie Marshes demand series, the downstream demand for the Macquarie was limited to a flow of 4,000 ML/d. This was the constraint under operation during the development of the Basin Plan, and has been carried through to the Northern

Page 66

Hydrologic Modelling for the Northern Basin Review

Basin Review. If the marshes have been pre-wetted, it is expected that these releases assist with passing low flows out the end of the marshes (i.e. a few hundred ML/d). The MDBA accepts there is some uncertainty surrounding the Macquarie model however for the reasons described Section 5.1 it was deemed fit for purpose when used as part of a comparative analysis where the baseline (i.e. pre-Basin Plan) model contained the same uncertainty. This view is supported by independent advice (Bewsher 2016). The MDBA also notes that a developing breakout near Mumblebone is likely to impact operational constraints, but this is unlikely to significantly affect the volumes requested in the demand series (see section 5.2.2).

Figure 20 Stylised hydrograph of catchment-scale (uncoordinated) demand series generation in the Macquarie Castlereagh Catchment. The shaded area represents the additional flow requested from storage. 5.2.4. Iterating a Model Scenario to Completion Water recovery results in a portion of the entitlement pool shifting towards an environmental water holder. In practice, the pattern of use for these entitlements will alter to achieve environmental outcomes, but they will still be subject to the same accounting and allocation system as irrigation entitlements. A core principle of the Basin Plan modelling approach is that the access rights of both irrigation and environmental water entitlements are to be maintained. This principle required a balance to be achieved between irrigation and environmental demands in the modelling platform. That is, in a ‘successful’ Basin Plan scenario, the inclusion of environmental demands will have no reliability impacts on other users in the system. The demand series process described above provided a 114-year pattern of environmental watering events to be requested in the models, subject to a model-external accounting mechanism. A result of the externality of the accounting mechanism was that it could not adapt to the daily accounting decisions (such as allocations, losses and carryover) and flow changes that occur dynamically in the model. The use of an annual environmental account volume outside of the model to generate environmental demands meant it was possible for the initial environmental demand pattern to under- or over-demand the recovered water. This Page 67

Hydrologic Modelling for the Northern Basin Review was tempered by the fact that the model was able to more accurately determine if these demands could be delivered. To further ensure that the environmental demands did not have any unintended impacts of third party reliability, the modelling process included an iterative component to refine the demand series based on model outputs — a feedback loop. As described in the Basin Plan hydrologic modelling report (MDBA 2012a), this iterative step had two core purposes:  to ensure that the scenario complied with the desired SDLs, and;  to ensure that there were minimal impacts on the reliability of other users in the modelled system. SDL compliance is a relatively straightforward test (i.e. a simple check of the long-term average diversion volume). However, without environmental accounts in the model to robustly monitor the water accounting process, measuring reliability impacts is not straightforward. The MDBA have therefore used the overall long-term storage behaviour and reliability of supply to ensure minimal reliability impacts, with the aim that they behave in a similar manner as in the baseline scenario. Any departure from this aim reflects modelling inaccuracies rather than policy intent. The iterative modelling approach made use of two levers to modify irrigation and environmental use in the model. Control of irrigation demands was provided by the general security ‘irrigable area’ parameter; control of environmental demands was provided by the ‘delivery ratio’ parameter. This delivery ratio applied to the downstream demand only (see section 4.2.3 in MDBA 2012a), and represents the additional losses borne by environmental entitlements to deliver water towards the end of each catchment. For each regulated catchment, multiple iterations were completed in which the irrigation and environmental demands were modified until the desired SDL, storage behaviour, and long- term reliability patterns were achieved. That is, the correct setting for the two levers was found by iterating against three components of model output data. To ensure meaningful consumptive entitlement reliability data, the irrigable area parameter was always held to be greater than or equal to the entitlement proportion. For example, if 10% of the general security entitlements had been recovered for the Basin Plan, then the irrigable area parameter reduction was no greater than 10%. This provided the desired results for the Namoi and Macquarie catchments. However, the Gwydir catchment contains significant levels of unregulated use that is less responsive to the irrigable area parameter, hence a third lever was added for this model to control the valley cap for general security off- allocation use. 5.2.5. Protection of Water in the Barwon–Darling Recovering water across the Northern Basin and using it for environmental flow delivery is expected to increase inflows to the Barwon–Darling, and this is reflected in the results of the Northern Basin Review. As a result of these increase inflows, Barwon–Darling users in the model had increased opportunities to extract water from the system. NSW have stated that the Water Sharing Plan includes an annual cap limit for licenced entitlement holders to ensure no growth in use, even under a scenario with increased inflows. This arrangement does not protect specific flow events from extraction. Instead, it will maintain the long-term average volume of users in the system, noting that their pattern of

Page 68

Hydrologic Modelling for the Northern Basin Review take could change in response to altered inflows. That is, this approach provides a ‘long- term volumetric’ protection of environmental water passing from upstream catchments through the Barwon–Darling. The water sharing plan version of the model is yet to experience an external audit, hence the MDBA have continued to use the available model for this catchment, which represents 2007/08 level of irrigation development and incorporates the Cap accounting rules of July 2007 (i.e. reduced entitlements and continuous carryover). This version of the model includes a capping mechanism but, unlike the water sharing plan version, it can display a growth in use in response to additional flows subject to hydrologic conditions and on-farm storage capacity — the model does not protect environmental water over the long-term. The MDBA therefore included a flow protection mechanism in the existing Barwon–Darling model by reducing the long-term annual extraction limit for all licenced irrigators. This mechanism was designed to approximate the long-term approach included in the updated Water Sharing Plan model. Independent reduction amounts were chosen for the three main reaches of the Barwon–Darling to ensure that, for each Basin Plan scenario, the distribution of diversions over the three reaches was the same as that displayed under baseline conditions (i.e. pro rata protection). NSW have advised that adopting this interim approach is an appropriate and pragmatic mechanism for limiting growth in use in the existing model given increased Barwon–Darling inflows. 5.3. Model Output Analysis The interpretation of modelling results should be viewed within context. Each model scenario represents only one possible set of outcomes resulting from the combination of a pre- determined water recovery pattern and an assumed targeted watering regime for the environment. The results are therefore indicative of a possible hydrologic regime that could occur — they are not a prediction of flows yet to occur. As described in the Basin Plan Modelling report (MDBA 2012a), there are a large number of methods that can be applied to analyse hydrologic data, and the MDBA has applied these on an as-needs basis. The Basin Plan is characteristically a long-term planning instrument, hence many of the statistics presented in this report are related to long-term analysis. A range of general hydrology and consumptive use results demonstrated the impacts of Basin Plan scenarios on water balances and flows through the river system. As a starting point, diversion data and flow through key indicator sites are often presented as long-term averages (such as an average volume of water past a site per year). Also, many of the SFIs (used to translate flow to environmental outcomes) consider the occurrence frequency of flow events of a specific shape. If an event had been requested through environmental demand series, the criteria for a successful event were relaxed by 10% in recognition of the uncertainty of the environmental demand series approach used in regulated catchments. A full explanation of the 10% allowance is given in Section 5 of the Basin Plan Hydrologic Modelling report (MDBA 2012a). Much of the Northern Basin experiences a highly variable climate, with a relatively wide spectrum lying between wet years and dry. For this reason, long-term averages provided only a starting point for the analysis — often of more interest (from environmental, social and economic perspectives) were the length of dry periods. The SFIs reported in Appendix A

Page 69

Hydrologic Modelling for the Northern Basin Review therefore also list the maximum dry periods that occur between the flow events under consideration. SFIs sample the hydrology at specific flows. The advantage of these indicators is that they have each been tied to a specific set of environmental outcomes. However, they often do not provide a complete picture of the changes in hydrology that occur between scenarios. As an alternative, flow duration curves display the occurrence frequency of all flows at a site. This document includes a number of differential flow duration curves (i.e. the percentage difference between two curves), which provide a straightforward method to compare model scenarios in terms of changes in flow frequency. These differential curves are ‘anchored’ by one or two core model scenarios — for example, a figure may be based on setting baseline conditions to be 0% (i.e. the starting point) and the 390 GL model scenario to be 100% (i.e. a fully implemented Basin Plan under its current settings), with all other scenarios represented as a change from this anchoring. The particulars of each differential curve are described throughout this document. An independent review of the modelling analysis methods confirms that they are based on long-established procedures and are technically sound (Bewsher 2016). Diversion and water recovery tables throughout this document are generally rounded to the nearest integer, hence the Northern Basin total recovery volumes may differ from a round- integer aggregate of all catchment-scale volumes.

Page 70

Hydrologic Modelling for the Northern Basin Review

6. Condamine–Balonne Scenarios The Condamine–Balonne region lies mainly in southern Queensland and extends about 100 km south-west into New South Wales. The major waterways in the region are the Condamine, Balonne and Maranoa rivers. A schematic map of the region, including structural features and flow constraints, is given in Figure 21. The Condamine River flows through southern Queensland and is only regulated in its upper reaches (by Leslie Dam near Warwick) and for a small section near Chinchilla. West of Surat, it becomes the Balonne River. The river passes through Beardmore Dam which, in conjunction with Jack Taylor Weir, provides water for the St George Irrigation Area and the township of St George. The Maranoa River joins the Balonne within the pondage area of Beardmore Dam. Downstream of Jack Taylor Weir the system breaks into a number of distributary channels with hydraulic characteristics similar to those displayed by a river delta. A small proportion of the water passes through the easternmost channel, the Narran River, which terminates in the Narran Lakes. But most flows pass through the complex distributary channels which form the Lower Balonne Floodplain, eventually providing water to the Barwon-Darling River. Nebine Creek irregularly contributes water to the Culgoa from the relatively flat Nebine catchment. From a river development perspective, the Condamine–Balonne is essentially an unregulated region with a relatively high level of extraction. As a proportion of inflows to the system, the public storage capacity is relatively low (234 GL), but the extraction is relatively high (42% under baseline conditions; the highest value in the Northern Basin). The entitlement framework downstream of Beardmore Dam is complex, comprising a number of different entitlement types related to various flow conditions. The St George Water Supply Scheme provides regulated water (against supplemented water licences) for users between Beardmore Dam and Jack Taylor Weir. However the majority of irrigation production relies on diverting unregulated flows into large privately owned off-stream storages, particularly downstream of St George — private storages comprise the bulk of storage capacity in the catchment, estimated to be 1,582 GL (Webb, McKeown and Associates Ltd 2007). Unregulated users often hold entitlements that allow take from multiple flow windows. Water is harvested from the river during low-to-mid flows; during higher flow events, when the river breaks out of the channel, water is extracted from overland flows. Water use in the Lower Balonne is therefore often categorised as either ‘water harvesting’ or ‘overland flow take’ (i.e. floodplain harvesting) depending on the river height measured at St George. Due to the natural morphology of the system (primarily the large distributary network forming the Lower Balonne Floodplain) the Condamine–Balonne has a relatively low level of connectivity downstream with the Barwon–Darling. The without development model, although not a precise representation of the natural system, provides an indication of the flow behaviour that would have occurred prior to development. It indicates that, under natural conditions, most of the water flowing into the system would have been consumed through natural processes such as evapotranspiration or seepage into groundwater systems, with approximately one third of flows passing downstream to the Barwon–Darling.

Page 71

Hydrologic Modelling for the Northern Basin Review

Development has decreased the connectivity of this system with the Barwon–Darling. A comparison of the baseline and without development models indicates that development has reduced the flow-through proportion from approximately 33% to 14%. These are long-term estimates — the actual flow-through proportion would vary event-by-event depending on the volume and peak flow of each event, and on the antecedent conditions. But the values are indicative of the changed flow conditions caused by on-river structures and consumptive use in the Condamine–Balonne. In 2012, the Basin Plan listed a recovery volume of 100 GL from the Condamine–Balonne catchment to achieve local environmental outcomes (with the allowance for additional recovery to contribute to the shared component and enhance downstream outcomes). The information base underlying this decision included the results from five model scenarios exploring recovery volumes between 60 and 203 GL (MDBA 2012a). A comprehensive modelling program has been completed for the Northern Basin Review to re-examine the local recovery volume. This comprised 24 model scenarios, divided into four scenario packages. For this purpose, the Condamine–Balonne was divided into the six sub- regions outlined in Figure 21, and described below.  Grey — Upstream Beardmore (tributary creeks) — these creeks lie away from the main stem of the river, and were not considered to be part of the water recovery process due to the relatively large conveyance losses to the Lower Balonne Floodplain and Narran Lakes.  Blue — Upstream Beardmore (main stem) — this region encompasses the main stem of the river in the Upper and Mid–Condamine, ending at Beardmore Dam.  Orange — St George — the ‘region’ in this case refers to all regulated users (i.e. supplemented entitlement holders) downstream of Beardmore Dam, receiving water through the St George Water Supply Scheme.  Red — Balonne River — encompassing unregulated users along the Balonne River between Jack Taylor Weir and Bifurcation 1 (where the river divides into the Culgoa and Balonne Minor Rivers).  Green — Narran system — this region encompasses users along the Balonne Minor between Bifurcations 1 and 2, and those along the Narran River; recovery along these two reaches has the greatest ability to enhance flows to the Narran Lakes.  Brown — Lower Balonne system — this encompasses this remaining parts of the catchment downstream of Bifurcation 1, and is defined by the main rivers and creeks of the Lower Balonne distributary network (the Culgoa, Bokhara, Ballandool, Birrie & Briarie). The first three scenario packages explored the effects of altering the recovery pattern, both geographically and by entitlement type. The design of these ‘recovery pattern’ scenarios was informed by conversations with the Northern Basin Advisory Committee and the Northern Basin Inter-Governmental Working Group. Their assistance is gratefully acknowledged. The environmental, social and economic findings from the recovery pattern scenarios were used to define a set of principles underlying the fourth package, exploring recovery volume options ranging from 65 to 176 GL. The findings from all scenarios were presented to the Authority to inform their decision regarding the SDL.

Page 72

Hydrologic Modelling for the Northern Basin Review

Figure 21: Schematic diagram for the Condamine–Balonne. The six sub-regions adopted for NBR modelling purposes are outlined in different colours.

Page 73

Hydrologic Modelling for the Northern Basin Review

6.1. Model Package 1 — Location of Recovery The overall purpose of these scenarios was to explore the general relationship between the geographic distribution of water recovery and the resulting flows through the system. Three of the scenarios explored the specific option of recovering water upstream of Beardmore Dam (balanced by reduced recovery downstream of St George). Also, drawing on a selection from the large number of available Condamine–Balonne scenarios allowed some general conclusions to be made regarding water recovery sensitivity. Of particular interest were the flows to the Narran Lakes, through the Lower Balonne Floodplain, and downstream to the Barwon–Darling. 6.1.1. Recovery Upstream of Beardmore Dam Three scenarios were completed for this study, representing recovery volumes upstream of Beardmore Dam of 0, 10 and 20 GL respectively. All scenarios represented essentially equivalent recovery volumes across the catchment (around 140 GL), but the water recovery was re-balanced between the Narran system (Figure 21) and the reaches upstream of Beardmore — that is, water recovery from the Lower Balonne Floodplain remained essentially constant between these scenarios. The recovery breakdowns for these scenarios are listed in Table 13. The model scenario identifiers are 1022 (0 GL), 1009 (10 GL), and 1010 (20 GL), where the numbers in parentheses refer to the volume recovered upstream of Beardmore. The recovery configurations are listed in Table 13. Water balance tables and SFI outcomes for these model scenarios are listed in Appendices A and C.

Page 74

Hydrologic Modelling for the Northern Basin Review

Table 13: Recovery pattern for the Upstream Beardmore scenarios Recovery Volume (GL) Entitlement Type Sub-Region Recovered in Model 0 GL 10 GL 20 GL (1022) (1009) (1010) Upstream Beardmore — 0 0 0 Dam (off main stem) Upstream Beardmore Unregulated 0 10 20 Dam (on main stem) Supplemented Medium St George 0 0 0 Priority Water harvester 9 9 10 (i.e. unsupplemented) JTW to B1 Overland flow 1 1 1 (i.e. floodplain harvester) Water harvester 29 19 15 (i.e. unsupplemented) Narran system Overland flow 12 11 7 (i.e. floodplain harvester) Water harvester 47 46 46 (i.e. unsupplemented) Lower Balonne Overland flow 43 44 44 (i.e. floodplain harvester) TOTAL 141 141 142

Table 14 lists the long-term average inflow to Beardmore for the three scenarios compared to baseline conditions. The 10 and 20 GL scenarios increased inflows to Beardmore by approximately 9 and 18 GL/y respectively compared to the 0 GL scenario. This indicates an approximate return of 90% related to water recovered upstream of the dam, with the remaining 10% lost to natural processes such as evaporation. This value is highly dependent on the location at which water is recovered — in the 10 and 20 GL scenarios, the water was recovered on the main stem of the river, of which approximately two-thirds was located just upstream of the Beardmore Dam near Surat (the other third was located towards the end of the Upper Condamine model, near Lemon Tree Weir). It is therefore likely the 90% return is located towards the upper end of possible values. Recovery of water on creeks away from the main stem of the river would return significantly lower values. Furthermore, existing water sharing arrangements in the Condamine–Balonne do not protect environmental water recovered upstream of the dam from extraction in the downstream reaches — this water would be subject to the same allocation arrangements as all other inflows to Beardmore. Testing through the models (pers. comm. DNRM) indicates that, under current arrangements, approximately one-third of additional environmental inflows would be extracted by downstream users. The model scenarios completed here have therefore included a long-term protection mechanism to ensure SDL compliance is maintained. The protection mechanism adopted for modelling purposes was to recover a small portion of additional entitlements downstream of the

Page 75

Hydrologic Modelling for the Northern Basin Review dam. This does not protect upstream environmental water on an event-by-event basis, but instead provides long-term volumetric protection. It is anticipated that, in practice, Queensland and CEWH officials would work towards an agreed protection mechanism that maintains reliability of all users in the system (including the environmental water holder) without the need for additional water recovery. Table 14 also lists the changes in flow at Brenda on the Culgoa River for the three scenarios. Overall, the 100 GL water recovery pattern downstream of St George (but excluding the Narran system) has increased flows at Brenda by an average of 69 GL/y (0 GL scenario; 1022). Supplementing this with 10 GL recovered upstream of the dam has provided an additional 4 GL/y at Brenda; 20 GL upstream has provided an additional 7 GL/y. Note that both of these values are subject to the protection mechanism described above. Also shown in Table 14 are the flows at Wilby Wilby for the three scenarios. The scenarios essentially represented a relocation of water recovery from the Narran system to upstream of Beardmore Dam, and this is reflected by the commensurate reduction in flow at Wilby Wilby. Overall, the options explored have increased flow through the Lower Balonne Floodplain balanced by a reduction of flow into the Narran Lakes.

Page 76

Hydrologic Modelling for the Northern Basin Review Table 14: Long-term average inflow to Beardmore, and flow at Brenda and Wilby Wilby under baseline conditions and for the ‘U/S Beardmore’ scenarios

Inflow to Beardmore (GL/y) Flow at Brenda (GL/y) Flow at Wilby Wilby (GL/y) Model Scenario (Run Number) Long-Term Change from Long-Term Change from Long-Term Change from Average Baseline Average Baseline Average Baseline Baseline 1065.0 — 223.7 — 75.3 — (845) 0 GL U/S Beardmore 1064.6 –0.4 292.8 +69.2 100.8 +25.4 (1022) 10 GL U/S Beardmore 1073.5 +8.5 296.9 +73.2 92.5 +17.2 (1009) 20 GL U/S Beardmore 1082.4 +17.4 299.7 +76.1 89.0 +13.6 (1010)

Page 77

Hydrologic Modelling for the Northern Basin Review

Whole-of-flow regime changes can be examined using the differential flow duration curve in Figure 22. This chart demonstrates that recovery upstream of Beardmore would provide increases over the whole flow regime, with the vast majority of the increase occurring in low-to- mid flows (i.e. less than 4,000 ML/d) — higher flows display only a marginal increase in frequency. Note that, under this scale, ‘200%’ indicates that the improvement in flow due to the Basin Plan has been doubled as a result of the altered spatial distribution of recovery. The without development scenario (not shown here) is situated far higher on this scale. The equivalent graph tracking flow changes at Wilby Wilby (Figure 23) indicates that the associated decrease in flow is spread fairly evenly over the whole flow regime through the Narran system.

Figure 22: Differential flow duration curve at Brenda for the U/S Beardmore scenarios. The anchor points in the chart are baseline flows (0%) and flows under the 0 GL U/S Beardmore scenario (100%).

Page 78

Hydrologic Modelling for the Northern Basin Review

Figure 23: Differential flow duration curve at Wilby Wilby for the U/S Beardmore scenarios. The anchor points in the chart are baseline flows (0%) and flows under the 0 GL U/S Beardmore scenario (100%).

Recovery Upstream of Beardmore Dam — Main Conclusions

 Current arrangements do not protect upstream water, and over the long-term approximately one-third would be extracted by downstream users.  If the water was protected, recovery upstream of Beardmore Dam would provide increased flows through the Lower Balonne, where the exact increase would depend on the location from which the water was recovered.  It is expected that most of the increased flow through the Lower Balonne provided by upstream of Beardmore recovery would be in-channel low-to-mid flows.

6.1.2. Flow to the Narran Lakes Downstream of St George, the Balonne River breaks into a number of distributary channels as part of the Lower Balonne Floodplain. The four major break points (bifurcations) are displayed in Figure 21. During low flow periods, the weirs at these bifurcations are used to manage flow downstream for environmental, stock and domestic purposes. Due to the natural morphology of the distributary network, only a small proportion of the flow at Bifurcation 1 reaches the Narran Lakes. Flow through the Balonne River divides into five separate channels through a number of bifurcations. The Culgoa and Narran rivers are the main channels (both carrying approximately one third of the flow over the long-term), but a significant proportion of flow through the Narran system is lost to natural processes such as evaporation, or is extracted for consumptive use prior to reaching the lakes. The NBR model scenarios were analysed to determine the relative contributions that could be made to the Narran Lakes through water recovery in different parts of the catchment. Specifically, the modelling was used to compare the relative ‘yields’ of water recovery in the

Page 79

Hydrologic Modelling for the Northern Basin Review

Narran system and recovery upstream of Bifurcation 1. Consistent with the SFIs, flows to Narran Lakes were measured at Wilby Wilby on the Narran River (422016). Eight scenarios were selected for this purpose, divided into three sets. Each set consisted of model scenarios with essentially equivalent water recovery volumes in the Narran system (17, 20 & 30 GL respectively), but with different recovery volumes upstream of Bifurcation 1. The characteristics of these scenarios are summarised in Table 15. These results are displayed graphically in Figure 24. The characteristics of the trend line are determined by the relative yields of water recovery in the two regions. The point where this line crosses the vertical axis indicates the yield from the Narran system alone (i.e. water recovery upstream of Bifurcation 1 equals zero), while the slope reveals the yield of upstream recovery. In this context, the three trend lines in Figure 24 have the same approximate slope, but have different vertical scales due to their relative Narran system recover volumes.

Table 15: Model scenarios used to examine the relationship between water recovery and flow to Narran Lakes

Water Recovery (GL) Wilby Wilby Flow (GL/y) Model Narran Set Narran Upstream Addition to Scenario Average System Bifurcation 1 Baseline Baseline 845 0 0 75 0 1112 17 37 91 15 1 (17 GL) 1040 11 75 93 18 1045 20 31 91 16 2 (20 GL) 1041 15 53 92 17 1089 20 52 94 18 1032 30 10 94 19 3 (30 GL) 1037 20 15 95 19 1109 30 64 97 22

Page 80

Hydrologic Modelling for the Northern Basin Review

Figure 24: Relationship between water recovery upstream of Bifurcation 1 and increase in flows to Narran Lakes Examining these trends indicates that, as a long-term average, approximately two-thirds of water recovered along the Narran system flows past Wilby Wilby and into the Narran Lakes. The value for each event would depend on antecedent conditions and the flow event characteristics. Also, the actual yield would depend somewhat on the location of recovery within the Narran system itself (i.e. water recovered along the Balonne Minor between Bifurcations 1 and 2 would display a lower value). But overall, the two-thirds rule is a useful guide. The slope of the lines in Figure 24 indicate that the relative yield of water recovered upstream of Bifurcation 1 is in the range 5 – 10% (again, as a long-term average). This proportion is highly dependent on the location of water recovery — due to natural transmission losses, water upstream of Beardmore Dam would provide a lower yield compared to water recovered downstream of St George.

Water Recovery and Flow to Narran Lakes — Main Conclusions

 Flows to Narran Lakes are best enhanced through water recovery along the Narran system  Due to the terminal nature of the Narran Lakes, water recovery along the Narran system will not pass downstream to the Barwon–Darling  Water recovery upstream of Bifurcation 1 can also provide flows to Narran Lakes, but with a substantially lower rate of return (around 5 to 10%)  The majority of water recovered upstream of Bifurcation 1 would be lost during transmission or pass through the other channels of the Lower Balonne Floodplain)

6.2. Model Package 2 — Entitlement Type Dependence The overall purpose of these scenarios was to explore the general relationship between the entitlement mix of the recovered water and the resulting flows through the system. This theme was divided into two distinct model scenario packages:

Page 81

Hydrologic Modelling for the Northern Basin Review

 Three scenarios exploring the mix of unregulated entitlements downstream of Jack Taylor Weir  A further scenario exploring the combined effects of regulated (i.e. supplemented) recovery from the St George Water Supply Scheme and an associated low flow release strategy from Beardmore Dam. Both packages are described separately below. 6.2.1. Unregulated Entitlement Mix Unregulated take in the Lower Balonne is typically broken into two categories: water harvesting, and overland flow take. As a general guide, water harvesting entitlements allow take from the river channel during low-to-mid flows, whereas overland flow entitlements are associated with extraction during overbank (i.e. mid-to-high) flow events. In this context, ‘low, mid, and high’ refers to the flow height measured at St George (not at the site of user extraction), which is used to define the allocations made to entitlement holders in the Lower Balonne. The river channel at this location has a substantial channel capacity, but it rapidly decreases downstream, especially when moving beyond the first bifurcation and into the Lower Balonne distributary network. It was therefore expected that the characteristics of flow enhancement achieved by the Basin Plan would be dependent on the mix of recovered entitlements. The purpose of this study was to quantify this dependence, and to determine the level to which the hydrologic differences would be reflected as changes in environmental outcomes (as measured by the SFIs). In essence, this study was exploring whether the anticipated environmental outcomes of the Basin Plan could be changed by altering the mix of recovered entitlements, but without altering the overall recovery volume in the Condamine–Balonne. This study was restricted to the unregulated (i.e. unsupplemented) entitlements downstream of Beardmore Dam. A separate study examining the possibility of recovering regulated (i.e. supplemented) water from users in the St George Water Supply Scheme to supplement low flows through the Lower Balonne is described in Section 6.4. Three scenarios were completed as part of this theme, summarised in Table 16. Each scenario supplemented the recovery that had been achieved at the time (Dec 2014; scenario 980 in Table 16) to achieve a total recovery volume of 140 GL in the Condamine–Balonne. Also, all scenarios had the same spatial breakdown of recovery. The difference between each scenario was the target entitlement type, summarised below:  Entitlement Scenario 1 (100% OLF) — all available OLF in each region was recovered, with the remainder made up from water harvester entitlements.  Entitlement Scenario 2 (80% OLF) — 80% of available OLF in each region was recovered, with the remainder made up from water harvester entitlements.  Entitlement Scenario 3 (water harvester focus) — remaining recovery focused on water harvester entitlements, supplemented by OLF recovery if required to meet the target total. The split between water harvester and overland flow entitlement recovery volumes for each scenario is displayed in Figure 25.

Page 82

Hydrologic Modelling for the Northern Basin Review

Table 16: Model parameters for the NBR entitlement type scenarios

Water Recovery (GL) By Scenario (Model run number) Region Entitlement Type Entitlement Option 1: Entitlement Option 2: Entitlement Option 3: Recovery at Dec 2014 Recover 100% OLF Recover 80% OLF Water harvester focus (980) (1046) (1048) (1047) Upstream Beardmore Dam 0 0 0 0 St George 0 0 0 0 Water harvester 15 15 20 34 JTW to B1 Overland flow 0 18 14 0 TOTAL 15 33 34 34 Water harvester 9 11 14 24 Narran System Overland flow 1 11 10 0 TOTAL 10 22 24 24 Water harvester 9 31 40 55 Lower Balonne Overland flow 12 53 42 27 TOTAL 21 84 82 82 Condamine–Balonne Total 47 140 140 140

Page 83

Hydrologic Modelling for the Northern Basin Review

Maximum All OLF recovered

Entitlement 1

Entitlement 2

Entitlement 3

0 20 40 60 80 100 120 140 Minimum Water Recovery (GL) No further OLF recovery OLF Recovery WH Recovery

Figure 25: Split between water harvester and overland flow recovery for the three entitlement scenarios Long-term flow results at Brenda (Culgoa River) and Wilby Wilby (Narran River) are listed for the three scenarios in Table 17. The long-term average flow at Brenda displays little-to-no dependence on recovered entitlement type — 116 GL of recovery throughout the floodplain and upstream of B1 has provide an additional 62 GL/y flow at this site. In contrast, flow at Narran displays more of a dependency, such that water harvester recovery increases the flow-through rate. This is largely an effect of the relative channel capacities of the Culgoa and Narran Rivers. The Narran River has a significantly lower capacity, hence higher flows through this channel are more susceptible to losses compared to the Culgoa River. Recovering OLF entitlements would preferentially restore higher flows, and these would therefore display greater losses (and hence a lower transmission efficiency) compared to water harvester recovery.

Page 84

Hydrologic Modelling for the Northern Basin Review

Table 17: Long-term average flows at Brenda (Culgoa River) and Wilby Wilby (Narran River) for the three entitlement scenarios

Flow at Brenda (GL/y) Flow at Wilby Wilby (GL/y) Model Scenario (Run Number) Long-Term Change from Long-Term Change from Average Baseline Average Baseline Baseline 223.7 — 73.5 — (845) Entitlement Scenario 1 100% OLF 285.7 +62.0 92.6 +17.2 (1046) Entitlement Scenario 2 80% OLF 285.3 +61.6 94.5 +19.1 (1048) Entitlement Scenario 3 Water harvester focus 285.2 +61.5 97.6 +22.3 (1047)

Figure 26: Differential flow duration curve at Brenda (Culgoa River) for the three entitlement scenarios. This diagram is anchored by baseline flow at 0%.

Page 85

Hydrologic Modelling for the Northern Basin Review

Figure 27: Differential flow duration curve at Wilby Wilby (Narran River) for the three entitlement scenarios. This diagram is anchored by baseline flow at 0%. Entitlement mix was found to have little impact on the long-term average flow at Brenda, however the types of flow enhanced by water recovery did display a dependency, as shown in Figure 26. Water harvesters extract low-to-mid flows (as measured at St George), hence recovering these entitlements (green line) would target these types of flow (i.e. those less than 8,000 ML/d at Brenda). Similarly, recovering overland flow users (blue line) would preferentially restore mid-to-high flow events (greater than 8,000 ML/d). The cross-over flow of 8,000 ML/d corresponds approximately to bankfull flow at this site. Most SFIs at this site correspond to bankfull-to-overbank flows, and recovering overland flow users would preferentially enhance the associated environmental outcomes for the riparian and floodplain zones (see the SFI results in Appendix A, Table A3). Compared to the in-channel metrics, the riparian and floodplain SFIs are associated with a larger surface area and a greater variety of ecological processes. A similar result is shown for the Narran River (Wilby Wilby; Figure 27), but with a crossover flow of around 4,700 ML/d. These flows are mostly in-channel, although some of the water will have entered breakout channels. The SFIs monitoring flow to Narran Lakes (Table A3) displayed the greatest improvement under scenario 1047 (water harvester focus). The Narran is a terminal lakes system with environmental outcomes that respond to a total volume of flow (rather than a threshold-based flow event), and the recovery of water harvester entitlements would preferentially restore those flow events with the lowest loss and highest frequency. For both sites, the modelling indicates that the mix of entitlements recovered in the Condamine– Balonne can be used to preferentially restore certain environmental outcomes. However, the differential flow duration curves (Figure 26 and Figure 27) demonstrate that this increase is counter-balanced by an associated decrease in other parts of the flow regime — there is a trade-off between flow types.

Page 86

Hydrologic Modelling for the Northern Basin Review

Entitlement Mix Scenarios — Main Conclusions

 The types of flow achieved depends on the mix of recovered entitlements  Recovering water harvesters would preferentially restore in-channel flows; recovering overland flow users would preferentially restore overbank flows. The cross-over point at both locations was 8,000 ML/d (Brenda) and 4,7000 ML/d (Wilby Wilby).  The flow changes were reflected in the environmental outcomes (SFI results in Appendix A). Lower Balonne Floodplain outcomes were best enhanced through overland flow recovery, while Narran Lakes outcomes were best enhanced through water harvester recovery.  Targeting certain entitlement types will preferentially restore certain flows, however this is balanced by a reduced response in other parts of the flow regime

6.2.2. Regulated Recovery Scenarios I (1103) & J (1115) included an environmental demand series associated with the recovery of 4 GL supplemented water from St George. The main purpose of these scenarios was to explore the sensitivity of outcomes to the SDL (see Section 6.4), however they also included environmental demand series to test whether a mechanism other than water recovery could influence low flows through this system. This demand was developed relatively late in the NBR modelling process, and was included because previous modelling indicated that water recovery would provide little change in low flows through the Lower Balonne, regardless of the volume recovered. This conclusion was based on the SFI results (Appendix A), but also on an analysis of baseflows that indicated shortfalls in the Lower Balonne occur disproportionality in drier periods (Appendix B). This presents a risk to Basin Plan outcomes, particularly in terms of waterhole persistence. An environmental demand series utilising supplemented water from the St George irrigation scheme was implemented to test one possible method of mitigating this risk. This demand series was included as a proof-of-concept test only. Additional iterations would be required to refine the demand to ensure it reflects a management approach that could plausibly be achieved in practice. Furthermore, these scenarios tested only one possible mechanism for enhancing low flow achievement (specifically through the recovery of regulated water with associated releases from Beardmore Dam). There are multiple options available to enhance low flows through the Lower Balonne, and additional work would be required to assess the relative merits of each approach. Under the option explored here, the environmental water holder would recover regulated water (4 GL in this case) and use this water to supplement existing environmental, stock and domestic (ESD) releases with additional environmental releases from Beardmore Dam. The demand series was created under the following principles:  Existing ESD events were supplemented if they had not achieved flow towards the lower end of the Culgoa River (measured at Weilmoringle).  Consistent with existing management practices, the demand series was capped to a flow of 830 ML/d (measured at Jack Taylor Weir) to ensure the flow did not pass above the pumping threshold of downstream consumptive users.

Page 87

Hydrologic Modelling for the Northern Basin Review

 The duration of the supplementary environmental releases was determined by the annual volume available in the environmental account (i.e. the EEST process; see Section 5.2). Overall, supplementary environmental releases were requested in approximately 80% of years, and the duration of these supplementary releases events were 6 days (or less, depending on water availability in that year).

Figure 28 Example supplemented low flow demand response Figure 28 shows an example model response to the supplement low flow demand. An additional five days of flow at 830 ML/d was demanded however the release was extend by only three days. This was a common result in the model scenario. Table 18 shows that about 17% (73 of 420) of the days of additional flow requested were delivered. Hence these additional flows only slightly modified the flow regime of the Lower Balonne without changing the achievement of flow indicators. The underlying cause for the reduced delivery was not clear at the time of writing this report. Further work would be required to ensure that the model is responding to the environmental watering pattern as desired, but also to ensure that the requested pattern is consistent with existing operating rules around Beardmore Dam as represented in the model.

Table 18 Condamine-Balonne regulated recovery response Number of occasions a Additional days of flow Additional days of flow supplementary demand was demanded achieved placed on existing ESD flows 98 420 73

Page 88

Hydrologic Modelling for the Northern Basin Review

Regulated Recovery — Main Conclusions

 The concept of environmental relases from Beardmore has been neither proven or disproven by Northen Basin Review Hydrological modleling  Additional investigations are required to detemine the practicality of implementing environmental releases from Beardmore Dam.  Additional investigations are required to detemine if regulated relaeses from Beardmore Dam can break critical dry spells at Weirmoringle on the Culgoa River and Wilby Wilby on the Narran River

6.3. Model Package 3 — Horizontal Slicing Unregulated users in the Lower Balonne often hold entitlements that allow take from multiple flow windows. ‘Horizontal slicing’ refers to the recovery of a segment of an entitlement (rather than a complete entitlement), where this segment is related to a specific flow window. This option could provide a more equitable spread of the water recovery (and hence the potential associated economic impacts) throughout the system. Also, horizontal slicing would allow specific flow windows — and therefore specific environmental outcomes — to be targeted by the recovery program. This targeted approach may not necessarily occur in practice. If the horizontal slicing option is pursued, it may result in a mix of recovered flow windows, in which case there will probably be little change in the resulting hydrology compared to standard whole-of-entitlement recovery. The purpose of the horizontal slicing scenarios was to investigate the capacity to preferentially restore certain flows through a targeted recovery program. To examine this option, Queensland modelling officials, in collaboration with the MDBA, completed three horizontal slicing model scenarios. This study was restricted to the unregulated (i.e. unsupplemented) entitlements downstream of Beardmore Dam, and included both water harvester and OLF extraction from the system. The first of these scenarios (1041) included standard whole-of-entitlement recovery of 97 GL; the remaining two scenarios replaced parts of this recovery with a highly targeted form of horizontal slicing (i.e. directed towards low flows or high flows; see the description in ‘In the 20 GL Low’ scenario, a few users displayed additional overland take from the system, resulting in a slightly reduced total recovery volume, however this is not expected to greatly impact the results. Table 19). The first scenario therefore acted as a central point of comparison, while the remaining two scenarios were completed as ‘bookend’ options to fully test the possible spectrum of results available to a horizontal slicing approach. Horizontal slicing was represented through the deactivation of allocations associated with a particular flow threshold. In the high flow option (scenario 1043), the highest flow slices were recovered to achieve 20 GL of recovery; similarly, the lowest 20 GL of slices were recovered for the low flow option (scenario 1044). Where required, the multi-year volumetric limit was imposed to ensure users did not display a growth in use through increased access in other parts of the hydrograph. This approach ensured SDL compliance of the partially recovery users over the long-term (noting however that take from individual events may be more or less than prior to the recovery).

Page 89

Hydrologic Modelling for the Northern Basin Review

The recovery distribution is shown in ‘In the 20 GL Low’ scenario, a few users displayed additional overland take from the system, resulting in a slightly reduced total recovery volume, however this is not expected to greatly impact the results. Table 19. The horizontal slicing option was explored in the JTW-to-B1 reach of the river only (Figure 21). All scenarios displayed the same recovery pattern from the Narran and Lower Balonne systems, and none of the scenarios included recovery upstream of Beardmore Dam or from supplemented users in the St George region. ‘In the 20 GL Low’ scenario, a few users displayed additional overland take from the system, resulting in a slightly reduced total recovery volume, however this is not expected to greatly impact the results.

Table 19: Recovery pattern for the horizontal slicing scenarios Recovery Volume (GL) Entitlement Type Sub-Region Recovered in Model Qld 100 GL 20 GL High 20 GL Low (1041) (1043) (1044) Upstream Beardmore — 0 0 0 Dam (off main stem) Upstream Beardmore Unregulated 0 0 0 Dam (on main stem) Supplemented Medium St George 0 0 0 Priority Water harvester 52 34 47 (i.e. unsupplemented) JTW to B1 Overland flow 0 17 0 (i.e. floodplain harvester) Water harvester 13 13 13 (i.e. unsupplemented) Narran system Overland flow 2 2 2 (i.e. floodplain harvester) Water harvester 12 12 12 (i.e. unsupplemented) Lower Balonne Overland flow 17 17 17 (i.e. floodplain harvester) TOTAL 97 96 92

The long-term average flows for these scenarios are listed in Table 20 for Brenda and Wilby Wilby. All three scenarios displayed approximately the same increase from baseline conditions at both sites, indicating that horizontal slicing would provide little change to the long-term average volume of water passing through these sites.

Page 90

Hydrologic Modelling for the Northern Basin Review

Table 20: Long-term average flows at Brenda (Culgoa River) and Wilby Wilby (Narran River) for the horizontal slicing scenarios

Flow at Brenda (GL/y) Flow at Wilby Wilby (GL/y) Model Scenario (Run Number) Long-Term Change from Long-Term Change from Average Baseline Average Baseline Baseline 223.7 — 75.3 — (845) Qld 100 GL Option 271.3 +47.6 92.1 +16.8 (1041) Horizontal Slicing 1: 20 GL High Flow 268.5 +44.8 90.4 +15.1 (1043) Horizontal Slicing 2: 20 GL Low Flow 271.2 +47.5 92.3 +17.0 (1044)

The differential flow duration curves for these scenarios are shown in Figure 29 (Brenda) and Figure 30 (Wilby Wilby), where the coloured lines trace the horizontal slicing scenarios and the dashed line traces the 100 GL option. On this scale, a value of 0% indicates no change from baseline flows; a value of 100% indicates that the flow frequency displayed by the Queensland 100 GL option (1041) was maintained. These results indicate that, although the long-term average flow did not significantly change, there was a change in the distribution of this volume over the flow regime. It can be seen that recovering certain flow windows preferentially restored the associated flows. Also, similar to the entitlement mix options, this increase was balanced by an associated decrease in the other parts of the flow regime. The change in hydrology showed only minor changes in the environmental outcomes for the Lower Balonne Floodplain and Narran Lakes (Table A3 in Appendix A).

Page 91

Hydrologic Modelling for the Northern Basin Review

Figure 29: Differential flow duration curve at Brenda for the horizontal slicing scenarios. The anchor points in the chart are baseline flows (0%) and flows under the Qld 100 GL option (100%; scenario 1041). Under both options, the volume subject to the horizontal slicing approach was 20% of the total recovery, and the deviation from the whole of entitlement scenario displayed in the Figure 29 and Figure 30 was generally around 20% over most of the flow regime. This indicates that horizontal slicing has a significant capacity to influence the frequency of specific flows achieved under the Basin Plan. However, the scenarios completed for this study represented the most highly targeted form of recovery available. It is unlikely that this level of targeting could be achieved in practice — it would require most (if not all) consumptive users to be willing to sell access in the desired flow window. It is therefore likely that recovery volume will be the primary driver of flow change through the system, and that horizontal slicing may be used as a refinement mechanism to favour the restoration of certain types of flow.

Page 92

Hydrologic Modelling for the Northern Basin Review

Figure 30: Differential flow duration curve at Wilby Wilby for the horizontal slicing scenarios. The anchor points in the chart are baseline flows (0%) and flows under the Qld 100 GL option (100%; scenario 1041).

Horizontal Scenarios — Main Conclusions

 Horizontal slicing could be used to preferentially restore certain types of flows  Targeting certain flow windows will preferentially restore certain flows, however this is balanced by a reduced response in other parts of the flow regime  Under the options considered, the flow changes were largely not significant enough to be reflected in the SFI results, indicating that preferentially restoring specific outcomes in the Lower Balonne Floodplain and Narran Lakes would require more than 20 GL of recovery to be targeted through horizontal slicing.

6.4. Model Package 4 — SDL Sensitivity Nine Condamine–Balonne scenarios were completed exploring SDL options ranging from 65 to 176 GL of recovery. Each of these nine scenarios were completed as part of whole-of-north model scenarios, described in Section 76. Figure 31 displays the recovery volumes divided by sub-region and entitlement type. These scenarios were analysed with the environmental, social and economic assessment frameworks, providing a substantial dataset to inform the Authority’s decision regarding the SDL. The design of these scenarios were informed by the findings from the recovery pattern scenarios described above. The following principles were thus adopted:

6 Scenarios B (1089) and H (1110) included equivalent setup parameters for the Condamine–Balonne. The only difference between these scenarios relates to the management of downstream water in the regulated catchments. These two scenarios are therefore discussed as a single item in this section, but discussed separately in Section 7.

Page 93

Hydrologic Modelling for the Northern Basin Review

 All scenarios included recovery upstream of Beardmore Dam (with the same environmental flow protection mechanism described in Section 6.1) to de-localise the potential economic impacts associated with water recovery.  Recovery from the Narran system was capped7 at 25 GL (based on both the environmental and economic assessment of previous scenarios).  Most scenarios included 4 GL of supplemented entitlement recovery from the St George Water Supply Scheme to provide the option for enhancing existing environmental, stock & domestic (ESD) releases (environmental water requirements for the Lower Balonne Floodplain and Narran Lakes emphasised the importance of breaking long cease-to-flow periods).  Based on the economic assessment, water recovery for the Lower Balonne Floodplain was divided in roughly equal proportions between the ‘St George to B1’ (yellow in Figure 31) and ‘Lower Balonne System’ (dark blue) sub-regions (see Figure 10 on page 34 for a schematic map of these sub-regions). Some of the scenarios also included the preferential recovery of specific entitlement types based on the environmental, social and economic assessment of prior model scenarios. Specifically, the 90 GL (scenario number 1112), 100 GL (1114) and 115 GL (1111 and 1115) scenarios included the preferential recovery of water harvesters along the Narran system to enhance Narran Lakes outcomes. Furthermore, the 100 GL scenario (1114) also included the preferential recovery of overland flow entitlements in the Lower Balonne system to enhance floodplain (both environmental and grazier) outcomes.

7 The only exception was the 176 GL range-finding scenario (1109) which included a 30 GL recovery from the Narran system to maintain recovery equity in the sub-regions downstream of the first bifurcation.

Page 94

Hydrologic Modelling for the Northern Basin Review

U/S Beardmore St George Supplemented St George to B1 — Water Harvesters St George to B1 — OLF & FPH

Narran — Water Harvesters Narran — OLF & FPH LBF — Water Harvesters LBF — OLF & FPH

1109

1108

1110

1089

1111

1115 Model ScenarioModel 1114

1103

1112

1113

0 20 40 60 80 100 120 140 160 180 200 Modelled Water Recovery (GL)

Figure 31: SDL options modelled for the Condamine–Balonne. The aggregated bar shows the total water recovered in the catchment, sub-divided by colour to represent sub-region and entitlement type recovered.

Page 95

Hydrologic Modelling for the Northern Basin Review Table 21: Long-term average flows through the Lower Balonne for the SDL sensitivity scenarios

Flow at Brenda (GL/y) Flow at Wilby Wilby (GL/y) Flow at Weilmoringle (GL/y) End of System Flow (GL/y) Recovery Volume (Run Number) Long-Term Change from Long-Term Change from Long-Term Change from Long-Term Change from Average Baseline Average Baseline Average Baseline Average Baseline Baseline 223.7 — 75.3 — 156.2 — 241.8 — (845) 65 GL 263.3 +39.7 82.4 +7.0 178.6 +22.5 269.1 +27.4 (1113; D) 90 GL 268.7 +45.1 90.5 +15.1 182.6 +26.5 276.4 +34.6 (1112; E) 100 GL 274.1 +50.4 91.8 +16.5 184.5 +28.3 280.0 +38.2 (1103; I) 100 GL 267.8 +44.1 94.4 +19.0 179.3 +23.2 278.5 +36.7 (1114; C) 115 GL 279.6 +55.9 95.9 +20.6 187.2 +31.0 284.6 +42.8 (1115; J) 115 GL 278.6 +54.9 96.7 +21.4 187.7 +31.5 284.1 +42.3 (1111; G) 142 GL 294.6 +70.9 93.5 +18.1 194.1 +37.9 293.4 +51.6 (1089 & 1110; B & H) 150 GL 296.8 +73.1 98.4 +23.0 195.1 +38.9 296.1 +54.3 (1108; A) 176 GL 301.4 +77.7 97.0 +21.7 197.1 +40.9 305.3 +63.5 (1109; F)

Page 96

Hydrologic Modelling for the Northern Basin Review

Figure 32: Differential flow duration curve at Brenda for the SDL sensitivity scenarios. The anchor point in this chart is baseline flows at 0%.

Page 97

Hydrologic Modelling for the Northern Basin Review

Figure 33: Differential flow duration curve at Wilby Wilby for the SDL sensitivity scenarios. The anchor point in this chart is baseline flows at 0%.

Page 98

Hydrologic Modelling for the Northern Basin Review

Figure 34: Differential flow duration curve for end of system flows under the Condamine–Balonne SDL sensitivity scenarios. The anchor point in this chart is baseline flows at 0%.

Page 99

Hydrologic Modelling for the Northern Basin Review

The long-term average flow for these scenarios at four key sites is listed in Table 21. Also shown are the differential flow duration curves for these scenarios at Brenda (Figure 32), Wilby Wilby (Figure 33) and for combined Condamine–Balonne end of system flows (Figure 34). The overall conclusion from these results is that increased water recovery provides increased flow through the system. The SFI results for these scenarios are provided in Appendix A with the environmental outcomes for each scenario presented and discussed by MDBA (2016c). The set included two scenarios representing 100 GL of recovery, but with different recovery strategies — Scenario C (1114) represented an entitlement-targeted recovery pattern to enhance Lower Balonne Floodplain and Narran Lakes environmental outcomes, whereas Scenario I (1103) followed a more generalised recovery pattern. The differential flow duration curve for Brenda (Figure 32) indicates that both scenarios displayed whole-of-flow regime increases, but the targeting affected the balance of this increase. Scenario C (purple line) showed a greater increase in higher flows compared to Scenario I (green line), but balanced by reduced low flows. The two lines cross at just under 10,000 ML/d (i.e. approximately bankfull flow), supporting the earlier findings that re-balancing the mix of water harvester and overland flow recovery results in a re-balance between in-channel and overbank flow outcomes. Similarly, the targeted recovery in Scenario C favoured lower flows through the Narran system compared to Scenario I (the purple and green lines respectively in Figure 33). Similarly, the two 115 GL scenarios represented similar recoveries, however Scenario G (1111) recovered more water upstream of Beardmore Dam (balanced by less recovery from the Lower Balonne system) compared to Scenario J (1115). Consistent with the findings from earlier scenarios, this has only slightly affected the long-term average flow through the system (Table 21), but there has been an impact on the types of flow. The differential flow duration curves (Figure 32 and Figure 34) indicate that Scenario G (orange line) has resulted in slightly less mid-to-high flows through the system compared to Scenario J (light blue line). That is, recovering water upstream of Beardmore Dam provides a slight reduction in mid-to-high flows, but a slight increase in low-to-mid flows. Overall, the differential flow duration curves indicate that recovery volume is the primary driver of increased flow through the system. It is possible to change the long-term average volume, or target a certain type of flow, but these are secondary ‘levers’ compared to recovery volume. An examination of Table 21 indicates that, on average, around one-third of the water recovered from the Condamine–Balonne will pass downstream stream. This trend is shown Figure 35. The precise value will depend on the distribution of recovered water, both geographically and by entitlement type, and on the protection of this water through the system, but the value of one- third is relatively stable. The corresponding trends at Brenda, Weilmoringle and Wilby Wilby are less stable, as they are far more dependent on recovery distribution, hence they are not discussed here.

Page 100

Hydrologic Modelling for the Northern Basin Review

70

60

50

40 System Flow (GL/y)Flow System

- 30

of - 20

10

0 Increased End Increased 0 50 100 150 200 Water Recovery (GL)

Figure 35: Relationship between water recovery volume and average increase in end-of-system flows for the Condamine–Balonne

Page 101

Hydrologic Modelling for the Northern Basin Review

7. Whole-of-north Scenarios The primary purpose of these scenarios was to examine the change in hydrology through the Barwon–Darling system resulting from different Basin Plan settings across the Northern Basin. As a result, most of the results presented in this section are for the Barwon–Darling. Associated results for the tributary systems from these scenarios are provided in the appendices. Also need to mention that the C–B setup for these scenarios was built using the findings from the C–B stand-alone scenarios As described in Section 4, the whole-of-north modelling work was divided into three distinct sets of model scenarios exploring, in turn, SDL range-finding, alternative policy settings, and refined scenarios. The first two sets of scenarios were completed to assist the Authority when choosing a set of options to be modelled in the third set. These model sets are described sequentially below. 7.1. SDL Sensitivity Four whole-of-north model scenarios were completed examining the sensitivity of flows through the Northern Basin as a function of recovery volume. The scenarios explored recovery volume options ranging from 278 GL (an estimate of existing recovery as of December 2015) to 415 GL. All of these scenarios were based on the same assumptions regarding future water recovery and water use:  Water Recovery — the existing recovery estimate adopted for the 278 GL scenario provided the starting point, and remaining recovery to achieve the desired reduction was achieved using the principles outlined in Section 5.1.  Water Use — water use was represented using the principles outlined in Section 5.2 — the whole-of-north coordinated approach (Strategy 1) was applied for the downstream watering component. This was the first package of whole-of-north model scenarios to be examined with the updated environmental, social and economic assessment framework provided for the Northern Basin Review. It was therefore designed to be a ‘range-finding’ package of scenarios, sampling a wide range of possible Northern Basin recovery volumes. A summary of the long-term average Barwon–Darling flows for these scenarios is presented in Table 22. For context, this table also includes the flows at these sites under without development conditions. The results indicate that water recovery across the Northern Basin provided an approximate linear increase in flows into and through the Barwon–Darling. Furthermore, the proportional ‘return’ (measured at Bourke) in these scenarios was around 50% — for example, 390 GL of recovery (plus water delivery) increased flow at Bourke by 197 GL/y. The return decreased a few percentage points upon reaching Menindee (mainly due to natural losses through the system, but also partly as a result of consumptive use downstream of Bourke). These values are subject to the underlying assumptions included in the modelling, but it is expected that they provide a good indication of the long-term yields that could be achieved in practice.

Page 102

Hydrologic Modelling for the Northern Basin Review

Figure 36 displays the differential flow duration curve for the increased flows at Bourke over the flow regime up to 80,000 ML/d. The data indicates that in-channel flows (i.e. less than around 25,000 ML/d) experienced the majority of the increase. Water recovery and use also increased the occurrence of overbank flows, but to a lesser extent.

Page 103

Hydrologic Modelling for the Northern Basin Review Table 22: Long-term average flow results for the five model scenarios testing SDL sensitivity

Long-Term Average Flow (GL/y) Inflow to B–D Bourke Louth Wilcannia Flow to Menindee Modelled Recovery (Run Identifier) Change Change Change Change Change Average from Average from Average from Average from Average from Baseline Baseline Baseline Baseline Baseline

Baseline 2770.6 — 2151.9 — 2072.9 — 1528.1 — 1723.2 — (845)

278 GL 2907.8 +137.2 2304.5 +152.6 2229.4 +156.5 1663.0 +134.9 1857.4 +134.2 (Scenario D; 1113)

320 GL 2925.8 +155.2 2320.1 +168.2 2243.5 +170.7 1677.4 +149.2 1871.6 +148.3 (Scenario E; 1112)

390 GL 2959.5 +188.9 2349.2 +197.3 2271.6 +198.7 1699.3 +171.2 1894.1 +170.9 (Scenario B; 1089)

415 GL 2971.3 +200.7 2358.9 +207.0 2280.4 +207.5 1707.6 +179.5 1902.8 +179.6 (Scenario A; 1108)

Without 4402.3 +1631.7 3811.6 +1659.6 3748.6 +1675.7 2818.8 +1290.6 3092.0 +1368.8 Development (844)

Page 104

Hydrologic Modelling for the Northern Basin Review

Figure 36: Differential flow duration curve at Bourke for the five SDL sensitivity scenarios. The anchor point in the chart is baseline flow (0%).

Page 105

Hydrologic Modelling for the Northern Basin Review

The differential flow duration curve provides a long-term change in flow frequency over the modelling period. To further investigate the changes on an event-by-event basis, the hydrographs at Bourke for the baseline and Scenario B (390 GL) were divided into a series of distinct flow events, and these events were matched between scenarios. This analysis provided an event-by-event comparison of changes in flow between each scenario. Figure 37 traces the peak flow for each event matched between the two scenarios, grouped into bins of width 1,000 ML/d.

100,000

90,000

80,000

70,000

60,000

50,000

40,000

30,000 Scenario B (390 GL) Peak Flow (ML/d)Flow Peak GL) (390 B Scenario 20,000

10,000

0

0

10,000 60,000 20,000 30,000 40,000 50,000 70,000 80,000 90,000 100,000 Baseline Peak Flow (ML/d)

Figure 37: Comparison between baseline and Scenario B (390 GL) of peaks of flow events at Bourke. The events have been divided into 1,000 ML/d increments. The dashed line traces a 1-to-1 relationship (i.e. no change from baseline flow). The results confirm that flow events in Scenario B (390 GL) displayed an increased peak compared to the same events under baseline conditions — fitting a trend line to the points in Figure 37 indicated an average increase in peak flow of 8% for each event. For in-channel events (i.e. <30,000 ML/d), the average increase was around 1,400 ML/d; for overbank events (>30,000 ML/d), the average increase was around 2,600 ML/d. The results for the four SDL sensitivity scenarios are summarised in Table 23.

Page 106

Hydrologic Modelling for the Northern Basin Review

Table 23: Average increased in peak flow for each event in the four SDL sensitivity scenarios Average Peak Flow Increase from Baseline (ML/d) Scenario (Recovery; GL) In-channel Events Overbank Events (<30,000 ML/d) (>30,000 ML/d) D (278) 1,042 1,870 E (320) 1,118 1,948 B (390) 1,383 2,571 A (415) 1,439 2,704

The SFIs provide further information regarding the changes occurring for identified ecologically important flow events. SFI results for these scenarios are summarised in Appendix A. A full description of the environmental outcomes associated with these scenarios can be found in MDBA (2016c), while the associated social and economic outcomes are described in MDBA (2012b). 7.2. Water Recovery and Use Sensitivity Three additional scenarios were completed to test the core assumptions and policy settings in the SDL sensitivity scenarios. Two of these scenarios tested the sensitivity of Barwon–Darling flow to the geographic location of recovery (i.e. the recovery pattern across the Northern catchments), while a third scenario tested the relationship with the assumed water delivery pattern. 7.2.1. Geographic Location of Recovery The two ‘recovery location’ scenarios had slightly different aims, and are therefore described separately below. Scenario C (350 GL) represented a near-equivalent recovery pattern to the 390 GL scenario, but with 40 GL less from the Condamine–Balonne. This scenario was therefore testing the sensitivity of Barwon–Darling flow to Condamine–Balonne recovery. The comparison scenario for this test was Scenario B (390 GL). In contrast, Scenario G (320 GL PR) represented a structural change to the overall recovery pattern. In this option, the principle of ‘future recovery will build on existing recovery’ (see the first principle in Section 5.1.1) was removed. Instead, the distribution of the shared recovery component followed a near-pure application of the default distribution described in the Basin Plan (i.e. a pro rata distribution based on the baseline diversion limit). This scenario was therefore testing whether a re-balancing of the Commonwealth portfolio to reflect a default distribution would affect flows through the Barwon–Darling. The comparison scenario for this test was Scenario E (320 GL; built from existing recovery). The long-term average flows for the two recovery location scenarios, and their respective comparison scenarios, are listed in Table 24.

Page 107

Hydrologic Modelling for the Northern Basin Review Table 24: Long-term average flow results for the model scenarios testing recovery location across the Northern Basin

Long-Term Average Flow (GL/y) Inflow to B–D Bourke Louth Wilcannia Flow to Menindee Modelled Recovery (Run Identifier) Change Change Change Change Change Average from Average from Average from Average from Average from Baseline Baseline Baseline Baseline Baseline

Baseline 2770.6 — 2151.9 — 2072.9 — 1528.1 — 1723.2 — (845)

320 GL 2925.8 +155.2 2320.1 +168.2 2243.5 +170.7 1677.4 +149.2 1871.6 +148.3 (Scenario E; 1112)

320 GL PR* 2934.9 +164.3 2311.7 +159.8 2231.4 +158.6 1664.2 +136.1 1858.6 +135.3 (Scenario G; 1111)

390 GL 2959.5 +188.9 2349.2 +197.3 2271.6 +198.7 1699.3 +171.2 1894.1 +170.9 (Scenario B; 1089)

350 GL 2943.0 +172.4 2333.0 +181.1 2255.9 +183.1 1686.2 +158 1880.7 +157.5 (Scenario C; 1114) *PR indicates a pro-rata method of water recovery

Page 108

Hydrologic Modelling for the Northern Basin Review

Figure 38: Differential flow duration curve at Bourke for the 350 GL and 390 GL scenarios. The anchor point in the chart is baseline flows (0%).

Page 109

Hydrologic Modelling for the Northern Basin Review

Scenario C (350 GL) — Reduction in Condamine–Balonne Contribution An examination of Table 24 indicates that, as expected, reduced recovery in the Condamine– Balonne (by 40 GL) resulted in a decrease in inflows (by 16.5 GL/y) and flow through the Barwon–Darling (by 16.2 GL/y at Bourke; 13.4 GL/y inflow to Menindee). This indicates that about 40% of the water recovered from the Condamine–Balonne is expected to flow into the Barwon–Darling system and through Bourke. The proportional ‘return’ would decrease as flow moves downstream and is subject to extraction and natural losses. The differential flow duration curve in Figure 38 indicates that this decrease was spread relatively evenly across the flow regime, indicating that the contribution of the Condamine– Balonne is important for all types of Barwon–Darling flow events. SFI outcomes for these scenarios are summarised in Appendix A. Overall, they indicate that the hydrology changes were not strong enough to trigger significant changes in SFI results for the Barwon–Darling. However, the changes in SFI results for the Condamine–Balonne itself resulting from the decreased recovery were significant. Scenario G (320 GL PR) — General Redistribution of the Shared Recovery Component A comparison of the catchment-scale recovery volumes for the two 320 GL scenarios is shown in Table 25 (note that the representation of water recovery at the sub-catchment scale followed the same principles in both scenarios). At the state scale, Scenario G has followed the default apportionment described in the Basin Plan. However, there has been a ‘rationalisation’ of the default approach at the catchment-scale — the Paroo, Warrego, Nebine & Moonie catchments were excluded from additional contribution to the shared recovery component. Hence, the contributions from the Condamine– Balonne and the Qld Border Rivers were slightly greater than the pure default values. Consistent with the overall approach for the Northern Basin Review (see section 5.1) the four smaller Queensland catchments were excluded from further recovery due to their relatively low ability to influence flows in the Barwon–Darling (the Moonie is well-connected downstream, but contributes relatively small volumes, whereas the remaining three catchments are poorly connected). Excluding these catchments from further recovery allowed the model scenario to be completed within NBR time frames. It is expected that, as the volumetric differences resulting from this rationalisation are small, it will not have a significant impact on the findings of this study.

Table 25: Differences in water recovery distribution for the two 320 GL options (note that all values have been rounded to integers). Grey rows indicate regions that were excluded from further contribution towards the shared recovery component. Green cells indicate catchments with an increase in recovery from Scenario E, orange indicates a decrease.

Water Recovery (GL) SDL Resource Unit Scenario E Scenario G Difference (320 GL) (320 GL PR)

Paroo 0 0 0 Warrego 8 8 0 Nebine 1 1 0 Moonie 2 2 0

QUEENSLAND

Page 110

Hydrologic Modelling for the Northern Basin Review

Condamine-Balonne 90 115 +24 Queensland Border Rivers 21 21 0 Queensland Total 123 147 +24 Intersecting Streams 8 0 –8

Gwydir 48 51 +3 NSW Border Rivers 7 13 +6 Namoi 20 20 0 Macquarie-Castlereagh 83 77 –6 Barwon-Darling 31 12 –19

NEW SOUTH WALES NSW Total 197 173 –24 Whole of North Total 320 320 0

Water recovery achieved to date is more advanced in NSW compared to Queensland, and the effects of this can be seen in Table 25. Moving from Scenario E (existing recovery + remaining default apportionment) to Scenario G (ignore existing recovery, near-pure default apportionment) shifted 24 GL of recovery from NSW to Queensland. The main changes occurred in the Barwon–Darling & Intersecting Streams (decreased recovery) and the Condamine–Balonne (increased recovery), with a few further exchanges between NSW catchments to balance the overall volume. In general, the redistribution from Scenario E to G shifted more of the water recovery into catchments with a lower level of connectivity to the Barwon–Darling. As a further note, the proportional split between regulated and unregulated entitlement recovery across the Northern Basin remained essentially constant between the two scenarios, hence both scenarios included the same volume of environmental water in public storages that was subject to the same operating principles. The modelling approach underlying environmental water use was therefore not a variable between the two scenarios, and the conclusions presented below can be directly tied to the redistribution and the relative downstream connectivity of each catchment. An examination of Table 24 indicates that the redistribution of recovery to purely a pro-rata recovery of entitlement between valleys increased inflows to the Barwon–Darling, but decreased flow through the sites along the river. This initially perplexing outcome is a direct result of the spatial redistribution — as expected, Barwon–Darling inflows increased as a result of shifting more recovery into the upstream catchments, however the increased inflows were more than offset by the upturn in consumptive use in the Barwon–Darling. These changes can be further explored through differential flow duration curves for both the Barwon–Darling inflows and flow at Bourke. Figure 39 indicates that the increased inflows to the system were spread over most of the flow regime, with the largest increases occurring for flows between 5,000 and 37,000 ML/d. Upon reaching Bourke, this pattern of increased inflows had reversed. The differential flow duration curve indicates that the increased Barwon–Darling extraction in Scenario G mainly

Page 111

Hydrologic Modelling for the Northern Basin Review impacted flow at Bourke in the range 0 – 15,000 ML/d. Flow events greater than this experienced a slight increase.

Page 112

Hydrologic Modelling for the Northern Basin Review

Figure 39: Differential flow duration curve for inflows to the Barwon–Darling for the two 320 GL scenarios. The anchor point in the chart is baseline flow (0%).

Page 113

Hydrologic Modelling for the Northern Basin Review

Figure 40: Differential flow duration curve at Bourke for the two 320 GL scenarios. The anchor point in the chart is baseline flow (0%).

Page 114

Hydrologic Modelling for the Northern Basin Review

Overall, the results indicate that the redistribution of water recovery pattern has reduced the frequency of flows <10,000 ML/d at Bourke, with a negligible change for the rest of the flow regime. The hydrologic changes are also reflected in the SFI results (Appendix A), indicating that environmental outcomes would also be affected by this redistribution. The results emphasise the importance of the location of water recovery — scenario G shifted some of the water recovery away from well-connected catchments (such the Intersecting Streams and Barwon–Darling) into less connected catchments (the Condamine–Balonne and Gwydir). As a general principle, recovering water in catchments with the greatest level of connectivity with the Barwon–Darling will provide the largest increase in flow through this system. Importantly, recovery in the Barwon–Darling itself provides the best ‘per-GL’ return for achieving increased flows through the Barwon–Darling system.

Redistribution of 320 GL Recovery — Main Conclusions

 The volume of water passing through the Barwon–Darling will be dependent on the location of any future water recovery across the Northern Basin.  Furthermore, recovery location can also affect the relative improvement experienced by different parts of the Barwon–Darling flow regime.  Overall flow through the Barwon–Darling is best improved through recovery from those catchments with the greatest level of connectivity with this system (notably, the Barwon–Darling itself).

7.2.2. Options for Delivering Water to the Barwon–Darling A single scenario was completed to elucidate the influence of the assumed coordinated downstream water delivery pattern present in all other scenarios. This scenario was constructed without coordinated releases from storages targeted at meeting site-specific flow indicators at Bourke. Instead, each catchment operated independently to supplement baseflows and enhance connectively to the Barwon-Darling. Table 26 summarises the main differences between the two watering strategies. Section 5.2.3 contains a detailed description of the two modelled alternate downstream environmental watering strategies. The two strategies encompass a relatively large range of the forecasting and coordination spectrum, hence they provide an indication of the range of possible flow improvement that could be achieved in the Barwon–Darling through regulated water releases in upstream catchments.

Page 115

Hydrologic Modelling for the Northern Basin Review

Table 26: Differences in assumed water use

Principle 390 GL Coordinated 390 GL Without Coordination Cross- Coordinated Not coordinated catchment regulated catchments jointly time their regulated catchments work individually coordination releases Downstream connectivity Barwon–Darling unregulated flow Release releases triggered to enhance and extend releases aim to supplement downstream trigger downstream connectivity to the Barwon- unregulated flows Darling Determinant Barwon–Darling SFIs Baseflows for release release pattern driven by 6,000 & 10,000 pattern informed by NSW Unregulated pattern ML/d at Bourke baseflows (B–D WSP) Upper limit for Bankfull & without development Baseflow, bankfull & without development target end of limit is the minimum of bankfull estimate & limit is the minimum of baseflows, bankfull system flows without development flows & without development flows

The changes in flow resulting from the altered watering strategy are shown in terms of long term averages in Table 27. Overall, the two 390 GL scenarios provided essentially equivalent long- term average flows into and through the Barwon–Darling. Figure 41 compares differential flow duration curves at Bourke resulting from the two modelled watering strategies. The 278 GL (D; 1113), 320 GL (E; 1112) and 390 GL (B; 1089) scenarios all included the coordinated strategy, whereas the blue line scenario (390 GL without coordination; scenario H; 1110) included catchment-scale watering only. This figure indicates that, although the two strategies provided the same long-term increase in flow, the distributions over the flow regime were significantly different. The results indicate that the without coordination scenario of water use preferentially increased the frequency of flows <5,000 ML/d at Bourke, offset by a smaller increase as higher flows. A further examination of the data indicated that this trend occurred consistently in dry, median and wet years. The results in Table 27 suggest that the future environmental release pattern from upstream storages will have little effect8 on the long-term average flows through the Barwon–Darling. But at a higher resolution, the results indicate that the adopted release pattern will influence the distribution of this increase both throughout time and across the flow regime.

8 This assumes a full utilisation of the held environmental water portfolio, an over-arching assumption of Basin Plan modelling.

Page 116

Hydrologic Modelling for the Northern Basin Review Table 27: Long-term average flow results for the model scenarios testing water delivery options across the Northern Basin

Long-Term Average Flow (GL/y) Inflow to B–D Bourke Louth Wilcannia Flow to Menindee Modelled Recovery (Run Identifier) Change Change Change Change Change Average from Average from Average from Average from Average from Baseline Baseline Baseline Baseline Baseline

Baseline 2770.6 — 2151.9 — 2072.9 — 1528.1 — 1723.2 — (845)

390 GL (with coordination) 2959.5 +188.9 2349.2 +197.3 2271.6 +198.7 1699.3 +171.2 1894.1 +170.9 (Scenario B; 1089)

390 GL (without coordination) 2962.2 +191.6 2351.2 +199.3 2274.8 +201.9 1700.0 +171.9 1894.8 +171.6 (Scenario H; 1110)

Page 117

Hydrologic Modelling for the Northern Basin Review

Figure 41: Differential flow duration curve at Bourke for the 278, 320 and both 390 GL scenarios. The anchor point in the chart is baseline flows (0%).

Page 118

Hydrologic Modelling for the Northern Basin Review

This conclusion is also reflected in the SFI results summarised in Appendix A. Under baseline conditions the 6,000 ML/d SFI at Bourke was achieved in 66% of years, and in 75% of years under the without coordination scenario (H; 1110). This indicates that the combination of water recovery and a low flow catchment-scale release pattern can improve environmental outcomes in the Barwon–Darling (largely through additional unregulated flows). Including a coordinated release pattern to target this type of flow provided additional improvement for this SFI (82% in scenario B; 1089). This indicates that a coordinated environmental release pattern is an alternative mechanism (i.e. beyond water recovery) that has the capacity to enhance the environmental outcomes achieved by the Basin Plan. The response for the two 10,000 ML/d SFIs at this site (also targeted by the coordinated watering strategy) was consistent with this conclusion, albeit with a lower strength. This indicates that the maximum flow that can be influenced through regulated releases is probably not much greater than 10,000 ML/d at Bourke.

Without Coordinated Delivery — Main Conclusions

 The long-term average flow into and through the Barwon–Darling system showed little dependence on the adopted environmental watering strategy.  There was also no change in long term average flow to Menindee.  However, a more refined examination indicates that the watering strategy has the capacity to influence which parts of the flow regime receive the greatest increase under the Basin Plan.  The low flow catchment-scale release pattern provided an enhanced set of outcomes for flows less than around 5,000 ML/d at Bourke.  Both watering strategies provided improved SFI results, indicating that the Basin Plan will improve environmental outcomes.  But the coordinated release pattern was found to enhance the targeted SFIs, indicating that a coordinated environmental release pattern has the capacity to enhance the environmental outcomes achieved by the Basin Plan given a set recovery volume.

7.3. Refined Scenarios Results from the scenarios described in Sections 7.1 and 7.2 were presented to the Authority for their consideration. Based on the hydrological findings, and the associated environmental, social and economic outcomes, two scenarios were developed incorporating a refined pattern of water recovery. The purpose of these scenarios was to provide a final set of information to the Authority as part of the review of Basin Plan settings in the North. Importantly, these are not ‘optimised scenarios’. Optimisation would require a stochastic approach — a large number of model scenarios each representing a specific recovery pattern, and assessed with the environmental, social and economic frameworks. Furthermore, this type of optimisation process would require a quantitative weighting of all outcomes to achieve a ‘best outcomes’ recovery pattern. Instead, the scenarios were built qualitatively using a set of judgements regarding the relative benefits of certain recovery patterns. These judgements are described for each scenario below. The Authority requested that these model scenarios represent Northern Basin water recovery volumes of 321 and 345 GL respectively. The Authority also selected a set of policy settings for

Page 119

Hydrologic Modelling for the Northern Basin Review these scenarios, specifically those relating to the future environmental water management practices and catchment-scale recovery volumes. The assumptions for future environmental water management closely followed those adopted for previous scenarios. The principles underlying demand series for local environmental water needs were unchanged, and Strategy 1 (whole-of-north coordinated releases) was adopted to manage the downstream water as this is most consistent with the overall ESLT approach (Section 5.2.3). The catchment recovery volumes for both scenarios are listed in the subsections below, along with the hydrological, environmental, and economic information underlying the selected volumes. 7.3.1. Refined 321 GL Scenario Table 28 lists the recovery volumes adopted for Scenario J (321 GL). The overall aim of this scenario was to explore the effects resulting from a recovery of 321 GL across the Northern Basin, but including a pattern of recovery informed by the assessment of previous scenarios. The qualitative judgements underlying this scenario are summarised in Table 28. As an example, the 20 GL adopted for the Namoi was based on a balance between the environmental and economic outcomes. That is, an environmental assessment of recovery options in the Namoi provided some confidence for SFI achievement with a recovery volume of 20 GL, while the economic assessment did not support recovery beyond this value. Also included in this judgement process were the hydrologic findings described in previous sections. The relative contribution of recovered water in each catchment towards flow in the Barwon–Darling was a key input, and these are summarised in Table 28. The only difference between the 320 GL Scenario E (1112) described in section 7.1 and this refined 321 GL scenario is the recovery pattern across the northern Basin. Overall, reduced recovery from the Macquarie–Castlereagh (from 83 to 55 GL) was balanced by increased recovery from the Condamine–Balonne (from 90 to 115 GL) and Barwon–Darling (from 31.5 to 36 GL) catchments, with a few minor changes in other catchments.

Page 120

Hydrologic Modelling for the Northern Basin Review Table 28: Catchment water recovery volumes for Scenario J (321 GL refined), and a summary of the contributing information from previous model scenarios Modelled Recovery Contributing Information from Assessment of Previous Scenarios (GL) SDL Resource Unit Summary 321 GL 320 GL Hydrological Environmental Economic Refined Paroo 0 0

Warrego 8 8 Unchanged from existing recovery Nebine 1 1 Moonie 2 2 Good improvement in SFI Recovery spread throughout Contributes large volumes (at 100 GL, plus 15 GL to outcomes at ~100 GL; recovery catchment to de-localise low-to-mid flows) downstream,

RECOVERY: 148 GL

maintain state share and includes 4 GL supplemented economic impacts (10 GL U/S Condamine-Balonne 90 115 but less frequently than NSW enhance downstream for low flow outcomes, and Beardmore, the remainder tributaries (due to climate flow targeted entitlement types for spread between St George and differences) Narran & LBF outcomes Dirranbandi) Local requirement (8 GL) Good downstream connectivity Around 40 GL required to

(42% OF(42% SHARED COMPONENT)

QUEENSLAND Queensland Border Recovery can be targeted away 21.4 21 plus 13 GL (state share & and good opportunities for achieve one of the three SFIs Rivers from Mungindi to reduce enhance D/S flow ) complementary measures to (dependent on recovery economic impacts NSW Border Rivers 7 7 Local requirement only enhance local env outcomes pattern)

Intersecting Streams 8 8 Unchanged from existing recovery Environmental benefit of Unchanged from existing Economic impacts limited to Gwydir 48 47 Low connectivity downstream further recovery is low due to recovery current level current operating constraints Additional recovery from Some confidence SFIs achieved Namoi 20 20 Regulated releases can Economic impacts at Wee Waa existing volume at recovery of ~20 GL enhance Barwon–Darling flows Macquarie- Reduction from existing SFIs can be achieved with Reduced economic impacts at 83 55 under certain conditions Castlereagh recovery reduced recovery Warren

NSW RECOVERY: 173 GL Unchanged from existing Barwon–Darling flows and environmental outcomes are best Economic impacts limited to

(58% OF(58% SHARED COMPONENT) Barwon-Darling 31.5 36 recovery enhanced through in-system recovery current level Whole of North Total 320 321 —

Page 121

Hydrologic Modelling for the Northern Basin Review Table 29: Long-term average flow results for the refined model scenarios

Long-Term Average Flow (GL/y) Inflow to B–D Bourke Louth Wilcannia Flow to Menindee Modelled Recovery (Run Identifier) Change Change Change Change Change Average from Average from Average from Average from Average from Baseline Baseline Baseline Baseline Baseline

Baseline 2770.6 — 2151.9 — 2072.9 — 1528.2 — 1723.2 — (845)

320 GL 2925.8 +155.2 2320.1 +168.2 2243.6 +170.7 1677.4 +149.2 1871.6 +148.4 (Scenario E; 1112)

321 GL 2924.4 +153.8 2321.7 +169.8 2246.0 +173.0 1679.7 +151.5 1873.9 +150.7 (Scenario J; 1115)

345 GL 2939.0 +168.4 2335.8 +183.9 2259.6 +186.7 1691.5 +163.3 1885.7 +162.5 (Scenario I; 1103)

Page 122

Hydrologic Modelling for the Northern Basin Review

(c)

(a)

(b)

Figure 42 Differential flow duration curves showing the change in inflows to the Barwon–Darling from (a) the Condamine–Balonne Bourke, (b) the Macquarie– Castlereagh, and (c) all tributaries combined. The anchor point in these charts is baseline flow (0%).

Page 123

Hydrologic Modelling for the Northern Basin Review

Figure 43: Differential flow duration curve at Bourke for the 320 GL and 321 GL (refined) scenarios. The anchor point in the chart is baseline flow (0%).

Page 124

Hydrologic Modelling for the Northern Basin Review

The change in inflows to the Barwon–Darling are shown as differential flow duration curves in Figure 42(a) for the Condamine–Balonne and (b) for the Macquarie–Castlereagh. The flows from each system have responded to the change in water recovery over essentially the whole flow regime. The total inflows to the Barwon–Darling (Figure 42(c)) display little change between scenarios, indicating little overall impact as a result of the shift in water recovery between tributaries. Table 29 shows a comparison of the long-term average flow results through the Barwon–Darling under this refined 321 GL scenario and the standard 320 GL recovery scenario (J; 1115). Overall, the refined recovery pattern has slightly increased flow through the Barwon–Darling. This is reflected in the differential flow duration curve shown in Figure 43, which indicates that the refined recovery pattern increased the frequency of flows greater than ~5,000 ML/d. This is somewhat offset by the reduced frequency of flows less than this threshold, but overall there was a net gain in flow through the Barwon–Darling. These improvements were reflected in the SFI results (Appendix A), which showed an increase in the frequency of 6,000 ML/d events at Louth and Wilcannia. There was also a general improvement in SFI outcomes for both the Lower Balonne Floodplain and Narran Lakes, largely due to an increase in recovery volume in the Condamine–Balonne, but also due to the targeted pattern of this recovery through this system and across entitlement types (see section 6). 7.3.2. Refined 345 GL Scenario The adopted recovery pattern for Scenario I (345 GL) is listed in Table 30. Similar to the 321 GL option, this scenario incorporated a number of judgements regarding the relative benefits of certain recovery patterns. The main changes made for this scenario compared to the 321 GL option were:  The total recovery volume across the Northern Basin was increased by 24 GL  The combined Queensland contribution remained unchanged, hence all of this increase was sourced from NSW catchments  The proportional contribution towards the shared component therefore tilted towards NSW  Within Queensland, the contribution to the shared component was sourced from the Moonie and Queensland Border Rivers (i.e. there was no contribution from the Condamine–Balonne) based on their relatively high level of downstream connectivity  Within NSW, the increased contribution to the shared component was sourced from the Namoi and Macquarie–Castlereagh catchments based on their ability to provide increased flow downstream during certain catchment conditions A comparison between the recovery patterns for the 321 and 345 GL is provided in Table 31. Similar to the 321 GL refined scenario, this 345 GL pattern of recovery does not represent an optimised option, it was instead based on a qualitative review of the findings from previous model scenarios.

Page 125

Hydrologic Modelling for the Northern Basin Review Table 30: Catchment water recovery volumes for Scenario I (345 GL refined), and a summary of the contributing information from previous model scenarios

Modelled Contributing Information from Assessment of Previous Scenarios SDL Resource Unit Recovery Summary (GL) Hydrological Environmental Economic Paroo 0 Warrego 8 Unchanged from existing recovery

Nebine 1 Added 2.5 GL to maintain Reduce economic impacts in Moonie 4.5 state share and enhance Good downstream connectivity the Condamine–Balonne & downstream outcomes Border Rivers Good improvement in SFI Recovery spread throughout

RECOVERY: 148 GL ~ 100 GL for local Contributes large volumes (at outcomes at 100 GL; recovery catchment to de-localise environmental outcomes, low-to-mid flows) downstream, includes 4 GL supplemented economic impacts (10 GL U/S Condamine-Balonne 100 no contribution to the but less frequently than NSW for low flow outcomes, and Beardmore, the remainder shared recovery tributaries (climate differences) targeted entitlement types for spread between St George and

2% OF2% SHARED COMPONENT) Narran & LBF outcomes Dirranbandi)

3

( QUEENSLAND Local requirement (8 GL) Good downstream connectivity Around 40 GL required to Recovery can be targeted away Queensland Border Rivers 35 plus 27 GL (state share & and good opportunities for achieve one of the three SFIs from Mungindi to reduce enhance D/S flow) complementary measures to (dependent on recovery economic impacts NSW Border Rivers 7 Local requirement only enhance local env outcomes pattern)

Intersecting Streams 8 Unchanged from existing recovery

GL Environmental benefit of Unchanged from existing Economic impacts limited to Gwydir 47 Low connectivity downstream further recovery is low due to recovery current level current operating constraints Additional recovery from Some confidence SFIs achieved Namoi 24 Regulated releases can Economic impacts at Wee Waa existing volume at recovery of ~20 GL

RECOVERY: 196 enhance Barwon–Darling flows Reduction from existing SFIs can be achieved with Macquarie-Castlereagh 74 under certain conditions Economic impacts at Warren recovery reduced recovery

NSW

8% OF8% SHARED COMPONENT) Unchanged from existing Barwon–Darling flows and environmental outcomes are best Economic impacts limited to

6

( Barwon-Darling 36 recovery enhanced through in-system recovery current level Whole of North Total 345 —

Page 126

Hydrologic Modelling for the Northern Basin Review

Table 31: Comparison of catchment-scale recovery volumes for the 321 and 345 GL refined scenarios Recovery Volume (GL) Catchment 321 GL Refined 345 GL Refined Change from 321 (1115) (1103) to 345 GL Paroo, Warrego, Nebine 9 9 0 Moonie 2 4.5 +2.5 Condamine–Balonne 115 100 –15 Queensland Border Rivers 21 35 +14 QUEENSLAND TOTAL 147 148 +1

NSW Border Rivers 7 7 0 Intersecting Streams 8 8 0 Gwydir 47 47 0 Namoi 20 24 +4 Macquarie–Castlereagh 55 74 +19 Barwon–Darling 36 36 0 NSW TOTAL 173 196 +23 TOTAL 321 345 +24

As a long-term average, the 345 GL scenario provided an additional 168 GL/y inflow to the Barwon–Darling, and an additional 184 GL/y flow through Bourke (Table 29). These values are in line with the expected values based on the trends observed in all other model scenarios (i.e. about a 50% flow-per-recovery ‘return’ measured at Bourke). The differential flow duration curves in Figure 44 show changes in inflows for the four main catchments that were modified between the 321 (green) and 345 GL (blue) refined scenarios (i.e. Condamine–Balonne, Border Rivers, Namoi and Macquarie–Castlereagh). Those catchments with additional recovery displayed an increase in inflow frequencies over most of the flow regime (flows <20,000 ML/d for the Border Rivers; < 14,000 ML/d for the Namoi; whole- of-flow regime for the Macquarie–Castlereagh). The Condamine–Balonne (less recovery) showed a decrease mainly in the <13,000 ML/d part of the flow regime. The resulting total change in Barwon–Darling inflow frequencies is shown in Figure 45. Most of the increase displayed under the 345 GL scenario (compared to 321 GL) was for inflow events <50,000 ML/d. A similar trend is shown at Bourke (Figure 46), which experienced an increase up to flows of around 45,000 ML/d.

Page 127

Hydrologic Modelling for the Northern Basin Review

(a) (c)

(b) (d)

Figure 44 Differential flow duration curves showing the change in inflows to the Barwon–Darling from (a) the Condamine–Balonne Bourke, (b) Border Rivers, (c) Namoi, and (d) Macquarie–Castlereagh. The anchor point in these charts is baseline flow (0%).

Page 128

Hydrologic Modelling for the Northern Basin Review

Figure 45: Differential flow duration curve for total Barwon–Darling inflows. The anchor point in the chart is baseline flow (0%).

Page 129

Hydrologic Modelling for the Northern Basin Review

Figure 46: Differential flow duration curve for total flows at Bourke. The anchor point in the chart is baseline flow (0%).

Page 130

Hydrologic Modelling for the Northern Basin Review

The increase in flow displayed by the 345 GL scenario were also reflected in the SFI results (Appendix A), specifically for the 6,000 ML/d indicators at Bourke, Louth and Wilcannia. Furthermore, the SFI results for the 345 GL scenario were essentially the same as the 390 GL scenario (B; 1089). Figure 46 shows the differential flow results for the 390 GL scenario (B; 1089), and this indicates that the 345 GL (refined) flow results progressed towards, but were still less than, the 390 GL results. However, the SFI results indicate that both the 345 and 390 GL scenarios have provided near-equivalent ecological responses throughout the Barwon–Darling. This emphasises the overall result that a targeted recovery strategy can be used to enhance the flow (and hence environmental) outcomes that can be achieved within a set recovery volume.

Page 131

Hydrologic Modelling for the Northern Basin Review

8. Alternate Mechanisms Analysis The Northern Basin Review model scenarios demonstrated that water recovery (combined with environmental water use) results in additional flow passing through the river system. However there are alternate mechanisms that can be used to achieve ecological outcomes. Throughout the term of the review, the MDBA received information regarding ‘complementary measures’ that could enhance individual flow events without permanent recovery of water for the environment. This section describes an investigation of the potential of these measures. There are a number of mechanisms that could be utilised to enhance individual flow events including (but not limited to):  One-off temporary trades  Options contracts over pumping opportunities (enduring agreements)  Utilisation of private storages (store and release).

The term event enhancement is used throughout this chapter to encapsulate all of the various mechanisms. Representing these measures directly in the modelling framework would be difficult and time- consuming. The models were built as water resource tools to inform water management, resource sharing, and storage operations. They do not include the capacity for dynamic economic interactions such as temporary trade or options contracts. Hence these measures cannot be implemented in the integrated hydrologic modelling framework (within the NBR timeframe) without introducing untenable uncertainty. Instead, a post processing analysis of model outputs was undertaken to identify opportunities to enhance individual events and hence enhance environmental outcomes through achieving more site-specific flow indicators. This analysis was therefore confined to the hydrological indicator site method (ESLT; MDBA 2011). The analysis identified the magnitude and frequency of hydrological opportunities to utilise event enhancement without considering the many other factors required for the various measures to be deemed practical (e.g. depth of the water market, capacity to trade and protect flow under current water sharing plans, and so on). The overall purpose of this analysis was to provide a first-pass test of feasibility for these alternate mechanisms. The analysis was conducted against the 390 GL recovery option (scenario B). It is expected that while another recovery option would provide somewhat different event enhancement values to those listed below, the overall feasibility conclusions would remain the same. Analysis was undertaken at eight of the indicator sites pertinent to the Northern Basin Review. These gauges are shown in Figure 47.The analysis does not consider the institutional and operational arrangements required to achieve the desired enhancement and, as it was performed at a point in space, the analysis does not consider flow attenuation. The opportunities for event enhancement are presented against the total pool of annual take in each catchment. This provides some additional water market context for the results, and was used to inform the overall feasibility judgement in each catchment. This is qualitative information only, and does not consider the water market apart in any other context apart from its absolute size, and does not account for any of the costs involved in implementing any of the potential event enhancement mechanisms.

Page 132

Hydrologic Modelling for the Northern Basin Review

Due to the limitations listed above the analysis does not prove or disprove the feasibility of event enhancement but rather can be used to guide further investigations by highlighting areas that have the best potential for event enhancement mechanisms to enhance environmental outcomes.

Figure 47 Stylised diagram of the Northern Basin, where the sites used for testing event enhancement opportunities are marked in light blue.

8.1. Methodology A systematic methodology was developed to detect opportunities (over the 114 year historical climate sequence) to enhance environmental outcomes through enhancing individual events. The analysis, termed the ‘Site-specific Flow Indicator Shortfall Analysis’ was designed to look for ‘near-miss’ SFI events and measure the additional volume required to turn them into successful SFI events. The SFIs are designed to provide a translation matrix between hydrology and environmental outcomes. Some SFIs are based on dry spells, while others are related to flow event of a specific volume, or a threshold and duration. Each required a specific version of the shortfall analysis, described in detail below. Table 32 lists the SFIs examined under this analysis, and provides each SFI with a measurement category

Page 133

Hydrologic Modelling for the Northern Basin Review

Table 32 SFIs analysed for event enhancement opportunities

Catchment Indicator Gauge Definition Categorisation 6,000 ML/d for 14 days Threshold/duration Bourke 10,000 ML/d for 14 days Threshold/duration 2 × 10,000 ML/d for 20 days Threshold/duration 6,000 ML/d for 20 days Threshold/duration Louth Barwon-Darling 21,000 ML/d for 20 days Threshold/duration 30,000 ML/d for 21 days Threshold/duration 30,000 ML/d for 30 days Threshold/duration Wilcannia 2 × 6,000 ML/d for 8 days Threshold/duration 20,000 ML/d for 5 days Threshold/duration 1,000 ML/d for 7 days Threshold/duration 3,500 ML for 14 days Threshold/duration Brenda 9,200 ML/d for 12 days Threshold/duration 15,000 ML/d for 10 days Threshold/duration 24,500 ML/d for 7 days Threshold/duration Condamine-Balonne 25 GL over 2 months Volume 50 GL over 3 months Volume Wilby Wilby 250 GL over 6 months Volume 100 GL over 12 months Volume 2 × 50 GL over 3 months Volume Weilmoringle Any flow >2 ML/d Cease-to-flow 500 ML/d for 75 days Threshold/duration Namoi Bugilbone 1,800 ML/d for 60 days Threshold/duration 4,000 ML/d for 45 days Threshold/duration 4,000 ML/d for 5 days Threshold/duration Border Rivers Mungindi 4,000 ML/d for 5 days Threshold/duration 2 × 4,000 ML/d for 11 days Threshold/duration

8.1.1. Threshold & Duration SFIs An SFI shortfall analysis for a flow rate based SFIs was performed at Bourke, Wilcannia, Brenda, Mungindi & Bugilbone as the SFIs at these indicator gauges are all defined in terms of a flow rate and a duration.  We looked for opportunities to enhance flow events and hence SFI achievement in the 390 GL option (scenario B).

Page 134

Hydrologic Modelling for the Northern Basin Review

 We only considered years when there was not a successful event in the baseline model scenario

 We only considered years where there was a successful event in the without development model scenario

See section 1.2 of this report for a more detailed description of the baseline & without development model scenarios.

Figure 48 Example SFI shortfall hydrograph Figure 48 shows an example of the shortfall analysis for a flow event passing through Brenda on the Culgoa River. The flow event is shown over a seven week period as it appeared under three model scenarios: without development conditions (blue), baseline conditions (orange) and in the 390 GL recovery scenario (grey). Also shown is the desired SFI event (dotted line). It can be seen that water recovery has increased the flow and duration of the event, however the 390 GL option has not satisfied the desired flow criteria. The green line traces the difference between the 390 GL event (grey line) and the SFI thresholds (dotted line). The area below the green dashed line represents the volumetric shortfall of the event. At each indicator site there were a number of SFIs that are each assessed individually.

Page 135

Hydrologic Modelling for the Northern Basin Review

8.1.2. Volumetric SFIs

Figure 49 Example Narran Lakes event Narran Lakes SFIs are associated with a volume of water entering the lakes (passing the Wilby Wilby gauge) within a certain time frame. As flow rates do not matter we can simply measure the size of flow events entering the Narran Lakes over the modelled sequence, any events not meeting the volumetric targets can simply be subtracted from the target to calculate the SFI shortfall. In the example event illustrated in Figure 3 the event was 23GL under the Basin Plan. The shortfall for the 25GL SFI for this event is therefore 2GL.

8.1.3. Dry Spell SFIs Environmental science conducted as part of the Northern Basin Review emphasised the importance of waterholes as refuges for fish populations during dry periods. This work identified the need to reduce the length of extended dry periods to maintain the viability of waterhole refuges. This (type of SFI) is defined by the cease-to-flow period. The gauge at Weilmoringle was used to define the dry spell SFI for the Culgoa River, and this SFI was analysed to identify opportunities to break dry spells through event enhancement. The four gauges along the Culgoa River used as part of this analysis are illustrated in Figure 50. The premise of the dry spell SFI shortfall analysis was to consider flows at all of these gauges and identify events in the Culgoa that ceased to flow before reaching Weilmoringle (the SFI site). The reduction in small flow-through events is largely an effect of Beardmore Dam which, although a relatively small public storage, absorbs and regulates the majority of upstream low flows. Part of this flow regime is maintained downstream through environmental, stock & domestic releases from Beardmore Dam, but the SFI and baseflow results indicates that shortfalls remain in the lower parts of the system.

Page 136

Hydrologic Modelling for the Northern Basin Review

Figure 50 Stylised diagram of the Condamine-Balonne Catchment Two distinct types of event were identified in the modelled time series . Figure 51 shows an example of an unregulated flow event and

Figure 52 shows an example of a regulated environmental, stock and domestic release from Beardmore Dam.

Page 137

Hydrologic Modelling for the Northern Basin Review

Figure 51 Example unregulated cease to flow event in the Culgoa River

Figure 52 Example regulated fresh in the Culgoa River An alternate mechanism would seek to enhance an existing event to provide additional flow through the system. The regulated releases were selected as being more appropriate for this type of enhancement for a number of reasons including:  consistent rates of rise and fall; and,  a lack of diversions as the flow rates are below pumping thresholds.

The analysis could then also be used to inform the potential of using regulated releases to meet SFIs in the Lower Balonne (see section 5.2.2).

Page 138

Hydrologic Modelling for the Northern Basin Review

8.2. Results and Discussion 8.2.1. Barwon-Darling

Figure 53: Bourke SFI shortfall analysis results Figure 53 shows the results of the shortfall analysis for SFIs at Bourke. Each column represents an opportunity to achieve an additional SFIs through event enhancement. The position on the horizontal axis shows at what time in the historical climate sequence the opportunity occurred, the height of the column denotes the additional volume required, and each colour represents one of the SFIs at the site. For example, the furthest left red columns represents an opportunity in the 1895-1896 water year to achieve SFI2 (10,000 ML/d for 14 days) with about 220 GL of event enhancement. That is, there was an event in the 390 GL recovery scenario that was about 220GL short of achieving SFI2. Table 33 displays the results of the same analysis, but configured to show the number of opportunities that become available as the capability to enhance individual events increases. An examination of Table 33 indicates that a 200 GL event enhancement capacity would provide 29 opportunities over the 114 year periods — that is, in about 25% of years. For context, long term average Barwon–Darling diversions in the 390 GL recovery scenario are about 165 GL/y. The results in Table 33 indicate that an alternate mechanism would therefore need access to a relatively large proportion of average Barwon–Darling diversions to provide an opportunity more than once a decade.

Page 139

Hydrologic Modelling for the Northern Basin Review

However, diversions in the Barwon–Darling display a relatively wide range, due to the highly variable climate (and hence river flows) through much of the northern Basin. Therefore filtering the number of opportunities by average diversions provides only a first indication of feasibility. Table 34 instead shows the results of the SFI shortfall analysis at Bourke as a proportion of annual diversion in the Barwon-Darling. This table lists the number of additional SFI events that could be achieved if a certain proportion of diversions were available to enhance events in any given year. Overall, the analysis suggests that event enhancement is not feasible for Bourke SFIs. Figure 53 displays the results of the SFI shortfall analysis at Louth. Again in the context of long term average diversions in the Barwon Darling of about 165 GL/y the analysis shows that event enhancement is also unfeasible for Louth SFIs.

Figure 54 Louth SFI shortfall analysis results Figure 54 displays the results of the SFI shortfall analysis at Wilcannia. Again in the context of long term average diversions in the Barwon Darling of about 165 GL/y the analysis shows that event enhancement is unfeasible for Wilcannia SFIs, however, there does appear to be a large number of opportunities to achieve SFI4.

Page 140

Hydrologic Modelling for the Northern Basin Review

Figure 55 Wilcannia SFI shortfall analysis results Wilcannia SFIs (Figure 55) appear to provide more event enhancement opportunities than other sites in the Barwon-Darling particularly for SFI4 (2 x 6,000 ML/d for 8 days). An event enhancement capacity of 50 GL would provide opportunities about 1 in 6 years. However, this volume is measured at Wilcannia, whereas most users in the system are located upstream of Bourke. There are significant loses between these two sites due to natural processes such as seepage and evaporation, and these losses are not included in the volumetric SFI shortfall estimates. It is highly likely that calculation factoring in these loses would show the volumes required to be unfeasible. It is conceivable that a complementary measure in an upstream tributary could enhance flow through the Barwon-Darling. The analysis was restricted to a consideration against diversions in the Barwon-Darling only because of both the significant losses associated with flows into the Barwon-Darling and the fact that existing water sharing arrangements do not allow for the protection of individual inflow events passing into this system from upstream tributaries.

Page 141

Hydrologic Modelling for the Northern Basin Review

Table 33 Barwon-Darling SFI Shortfall Analysis Results

Number of Opportunities (over 114 years) Site SFI given Diversions Available (GL) 5 10 50 100 200 400 SFI1 6,000 ML/d 3 3 8 20 21 21 (14 days)

SFI2 10,000 ML/d 0 0 0 7 11 33 (14 days) Bourke SFI3 10,000 ML/d 0 0 0 0 0 18 (2 × 20 days)

TOTAL 3 3 8 27 32 72

SFI1 6,000 ML/d 0 0 4 24 37 37 (20 days)

SFI2 21,000 ML/d Louth 0 1 4 10 18 25 (20 days)

TOTAL 0 1 8 34 55 62

SFI1 6,000 ML/d 4 7 27 47 60 60 (2 × 7 days)

SFI2 20,000 ML/d Wilcannia 3 3 8 15 28 32 (7 days)

TOTAL 15 20 49 78 111 118

Page 142

Hydrologic Modelling for the Northern Basin Review

Table 34 Barwon-Darling SFI Shortfall Analysis Results in terms of Annual Diversions

Number of Opportunities (over 114 years) Site SFI given Proportion of Diversions Available 5% 10% 20% 50% SFI1 6,000 ML/d 2 2 4 15 (14 days)

SFI2 10,000 ML/d 0 0 0 1 (14 days) Bourke SFI3 10,000 ML/d 0 0 0 0 (2 × 20 days)

TOTAL 2 2 4 16

SFI1 6,000 ML/d 0 1 2 17 (20 days)

SFI2 21,000 ML/d Louth 1 1 4 10 (20 days)

TOTAL 1 2 6 27

SFI1 6,000 ML/d 6 9 20 39 (2 × 7 days)

SFI2 20,000 ML/d Wilcannia 3 3 4 11 (7 days)

TOTAL 18 25 38 65

8.2.2. Condamine-Balonne The Lower Balonne contains flow threshold based, volumetric and dry spell SFIs; the results for each are detailed below.

Page 143

Hydrologic Modelling for the Northern Basin Review

Flow Threshold SFIs

Figure 56: Brenda SFI shortfall analysis results Figure 56 and Table 35 display the results of the SFI shortfall analysis performed at the Brenda gauge. The same results are shown in Figure 57 and Table 36, but in terms of annual diversions in the Lower Balonne.

Table 35 Brenda SFI shortfall analysis results

Event Enhancement SFI1 SFI2 SFI3 SFI4 SFI5 Total Capability (GL)

5 1 0 0 0 0 1 10 2 1 0 0 1 4 50 26 22 28 8 3 87 100 26 24 36 23 9 118 200 26 24 36 28 15 129

Page 144

Hydrologic Modelling for the Northern Basin Review

Figure 57 Brenda SFI shortfall analysis results with annual Lower Balonne diversions Table 36 Brenda results as a proportion of annual diversions

PROPORTION OF OPPORTUNITIES DIVERSIONS AVAILABLE SFI1 SFI2 SFI3 SFI4 SFI5 Total 5% 0 1 0 0 0 1 10% 0 2 5 2 1 10 20% 1 4 18 8 4 35 50% 2 11 29 22 12 76

While Figure 56 and Table 35 appear to show a reasonable number of opportunities to enhance individual flow event to achieve more SFIs at Brenda, many of these opportunities occur in years when Lower Balonne diversions are significantly below the average. The significantly lower number of opportunities in Table 37 indicate the many of these opportunities (particularly for SFI1) cannot be acted upon without acquiring a significant proportion of the diversion entitlements. The most prospective type of flow that could be further explored are bankfull flow events, measured through SFI3 (9,200 ML/d). The capacity to access 20% of upstream take provided 18 such opportunities over the 114 years (or, approximately once every six years over the long-term).

Page 145

Hydrologic Modelling for the Northern Basin Review

Dry Spell SFIs 116 regulated stock and domestic releases were detected in the Culgoa River in the Northern Standard model run. Each series in Figure 58 represents a single event with each point representing the volume of the event past each gauge. As shown below in Table 37, 49 of those events resulted in a flow at Weilmoringle and 67 did not. Table 37 also shows that on average the volume at Whyenbah of events that did not make it to Weilmoringle were larger than those that did.

Figure 58 Culgoa freshes analysis results This result is counterintuitive as it is reasonable to suspect that larger events are more likely to result in a flow at Weilmoringle. The result indicates that the volume required to produce a flow through event in the Culgoa River is highly uncertain. Antecedent catchment conditions are likely the dominant factor in whether an event reaches Weilmoringle rather than the volume of the event itself.

Table 37 Culgoa regulated freshes analysis results

SFI success SFI Fail Count 49 67 Average Volume at Whyenbah (GL) 2.1 4.7

Although the volume required is uncertain the range of values (3 to 15GL at Whyenbah) means that in the context of average Lower Balonne diversions of 117 GL/y it is feasible that flow through events could be generated with event enhancement mechanisms. Further investigation is required to determine if these events could be generated during extended and ecologically significant dry spells.

Page 146

Hydrologic Modelling for the Northern Basin Review

Volumetric SFIs

Figure 59 Narran Lakes SFI shortfall analysis results As shown in Figure 59 and Table 38 there is a steady increase in event enhancement opportunity as the volume available increases. The majority of these opportunities relate to the smaller of the SFIs at Wilby Wilby (25 & 50 GL events).

Table 38 Narran Lakes SFI shortfall analysis results

Event Enhancement Number of Opportunities Frequency of Opportunities Capacity (GL) (over 114 years) (proportion yrs) 0 0 0% 2 4 4% 4 9 8% 6 15 13% 8 21 18% 10 27 24% 12 35 31% 14 43 38% 16 50 44% 18 58 51% 20 68 60%

Page 147

Hydrologic Modelling for the Northern Basin Review

8.2.3. Namoi

Figure 60 Bugilbone SFI shortfall analysis results The results displayed in Figure 60 and Table 39 suggest that additional Namoi SFI events could be generated roughly 1 in 4 years with an enhancement capacity of 20GL. This conclusion is supported by the results shown in Figure 61 and Table 40, which have further filtered the opportunities based on annual diversions. This is a positive outcome, but further analysis would need to consider issues pertinent to a regulated catchment such as the Namoi such as operating rules and channel sharing with irrigation delivery.

Table 39 Bugilbone SFI shortfall analysis results

Event Enhancement SFI 1 SFI 2 SFI 3 Total Capacity (GL)

5 5 0 1 6 10 8 1 1 10 20 23 6 13 42 50 31 31 11 73 100 31 31 12 74

Page 148

Hydrologic Modelling for the Northern Basin Review

Figure 61 Bugilbone SFI shortfall analysis results with annual Namoi diversions

Table 40 Bugilbone results as a proportion of annual Namoi diversions

PROPORTION OF DIVERSIONS OPPORTUNITIES AVAILABLE SFI1 SFI2 SFI3 Total 5% 10 4 1 15 10% 23 8 6 37 20% 31 18 10 59 50% 31 20 12 63

Page 149

Hydrologic Modelling for the Northern Basin Review

8.2.4. Border Rivers

Figure 62: Mungindi SFI shortfall analysis results The SFI shortfall analysis results in the Border Rivers (shown in Figure 62) indicate there are hydrologic opportunities to generate additional events for all three SFIs at Mungindi. Table 41 shows the number of opportunities increase markedly between an event enhancement capacity of 20 and 40 GL. This conclusion is supported by Table 42 and Figure 63, which display the results of the SFI shortfall analysis at Mungindi in terms of annual diversion in the Border Rivers. Access to temporarily acquire 10% of diversions on the Border Rivers would result in opportunities to achieve SFIs about 1 in 3 years. Overall, there is potential in the Border Rivers to use regulated releases in conjunction with other event enhancement mechanisms in order to enhance environmental outcomes.

Table 41: Mungindi SFI shortfall analysis results

Event Enhancement SFI 1 SFI 2 SFI 3 Total Capacity (GL) 5 0 0 2 2 10 0 0 4 4 20 2 4 8 14 40 14 25 13 52 60 19 40 20 79 80 19 40 26 85 100 19 40 28 87

Page 150

Hydrologic Modelling for the Northern Basin Review

Figure 63 Mungindi SFI shortfall analysis results with annual Border Rivers diversions

Table 42 Border Rivers results as a proportion of annual diversions

PROPORTION OF DIVERSIONS OPPORTUNITIES AVAILABLE SFI1 SFI2 SFI3 Total 5% 1 1 7 9 10% 11 16 14 41 20% 19 37 74 130 50% 19 40 80 139

8.3. Alternate Mechanisms Conclusions Analysis of model outputs from the NBR hydrologic modelling has shown that there is potential for the achievement of more SFIs through enhancing individual events, however the opportunities to do so are not distributed evenly among the various catchments of the Northern Basin. The volume of water required and the infrequency of opportunities make event enhancement unfeasible at the three indicator sites along the Barwon-Darling. In the Lower Balonne the analysis has shown there is potential to increase the achievement of SFIs related to the Narran Lakes and to dry spell indicators in the Culgoa and Narran Rivers. In the Namoi and the Border Rivers access the about 10% of the consumptive pool would provide opportunities to create additional SFI events roughly 1 in 6 and 1 in 3 years respectively. The feasibility of alternate mechanisms to water recovery will be determined by more than hydrology alone. In addition to hydrologic opportunities, new institutional and operation arrangements would be required. A consideration of the financial costs and the markets which these mechanisms will operate would also be necessary. The hydrologic analysis undertaken as

Page 151

Hydrologic Modelling for the Northern Basin Review part of the NBR do not prove nor disprove the feasibility of these mechanisms but rather can be used to guide and refine future investigations.

Alternate Mechanisms — Main Conclusions

 Hydrologic opportunities exist to enhance environmental outcomes through enhanced individual flow events.  The Lower Balonne shows the best potential for enhanced environmentl outcomes through alterative event enhancement mechanisms. This relates to breaking dry spells in lower reaches of the regions rivers and creaing environmentlaly significant events in the Narran Lakes.  Alternate mechanisms also show potential to enhance enironmnetal outcomes in the Border Rivers and the Namoi.  Much more inverstigation and consideration is required before any such mechanism can be deemed practical.

Page 152

Hydrologic Modelling for the Northern Basin Review

9. Downstream Deliverability Analysis The Basin Plan includes an element of water recovery to be shared across the Northern Basin with the aim of enhancing outcomes in the Barwon–Darling. In the modelling scenarios it was therefore assumed that, in the future, this component of water will be used in line with Basin Plan aims to achieve additional flows through the Barwon–Darling. Representing a future pattern of water use is difficult because there is very little practical experience on this front. Downstream environmental water use in the Northern Basin has occurred in only a few exploratory trials to date. For this reason, a number of detailed assumptions were made regarding downstream water use included in the NBR modelling (section 5.2). It is acknowledged that making regulated releases to enhance existing flows through the Barwon–Darling contains a unique set of challenges (section 5.2.3). For this reason, two different strategies were modelled — together encompassing a broad range across the ‘forecasting and coordination spectrum’ — to quantify the effects of these assumptions. A comparison of the outcomes from these two scenarios was presented in section 7.2.2. Of the two modelled options, Strategy 1 (whole-of-north coordinated releases; section 5.2) represents a greater change to existing operational practices and requires a greater capacity to forecast flows than is currently available. To further investigate this downstream watering approach, this section describes a bespoke analysis, designed to more fully understand the dynamics of environmental water use in the modelling framework. This analysis was undertaken using the 390 GL scenario with coordinated downstream demands (i.e. scenario B; 1089). The environmental watering included in this scenario followed the standard Basin Plan modelling process, in which demands were generated for both in-valley and downstream (i.e. Barwon-Darling) environmental outcomes, and downstream demands were placed at the most downstream suitable model nodes available. In an attempt to satisfy environmental water requirements in the Barwon-Darling, water was released from storages in the regulated and semi-regulated catchments (shown in Figure 64) to enhance unregulated events. Essentially the demand series represents the environmental water holder prioritisation process while the model represents the application of these priorities by river operators. See section 5.2 for a more detailed description of this process.

Page 153

Hydrologic Modelling for the Northern Basin Review

Figure 64: Stylised representation of northern Basin catchments The analysis below is divided into two parts. The first part examines the downstream demand series to determine the additional flow being requested for delivery (i.e. characteristics of the water orders). The second part examines the output from the models (i.e. the achieved downstream water delivery). 9.1. Demand Series Analysis The environmental watering strategy constructed outside of the modelling framework was based on environmental water requirements, climatic conditions and environmental water availability. The demand series were then placed within the regulated catchment models that applied further practical constraints such as channel and storage release capacities while attempting to supply the demand series. Downstream demands were activated in 40% of years, during which coordinated releases were requested from four catchments (Border Rivers, Gwydir, Namoi & Macquarie) to supplement

Page 154

Hydrologic Modelling for the Northern Basin Review existing flows through the Barwon–Darling. The period of coordinated releases (i.e. the downstream demand window) lasted for around four weeks for each event.

Figure 65 Summary of downstream demands Importantly, these releases supplemented existing flows in each tributary — that is, the demand series were not seeking to build end of system flows from scratch. Figure 65 shows a long-term average summary of the demands placed in the regulated catchment models. The blue bar represents the average pre-existing (i.e. baseline) flow at the downstream demand site during the demand window, and the orange bar shows the average requested additional flow requested through environmental releases from the upstream storage. On average, around 75% of the end of system flow from each tributary was provided by pre-existing flow, while the remaining 25% was to be sourced from storage. Note that this applies to the demand series only. Analysis was undertaken to better understand how these demands translated into additional flow once they have been included in the models. 9.2. Downstream Water Delivery To analyse the modelled response to downstream demands, two sets of analysis were undertaken. The first focused on the change in end of systems flows due to environmental demands (demand window analysis), and the second focused on the ability of these releases to meet environmental water requirements in the Barwon–Darling (event success analysis). The method and results of this work is described separately below. 9.2.1. Demand window analysis In order to analyse the modelling response to downstream environmental demands over the entire 114 year historical climate sequence, a systematic method of assessing the response to each demand needed to be developed. The concept is based on a comparison of Baseline and Basin Plan model (see section 2 for a detailed description of these scenarios) end of system flow time series for a 30 day window centred on the demand series.

Page 155

Hydrologic Modelling for the Northern Basin Review

Figure 66 Stylised representation of the Namoi River Catchment

Figure 67 Example demand response from the Namoi model The demand window analysis was performed on inflows to the Barwon-Darling at the end of each regulated catchment rather than at the demand model node. For example in the Namoi catchment the downstream demands were placed at the Goangra gauge (see Figure 66) in order to supply flows to the Barwon River, but the analysis was performed at the end of system

Page 156

Hydrologic Modelling for the Northern Basin Review model node downstream of the Namoi-Pian junction. This ensured that the analysis was revealing all changes in end of system flows rather than measuring only the subset of changes achieved at the demand site. As an example, Figure 67shows the change in end of system flows in the Namoi as a result of water recovery and water delivery, including that requested by the downstream demand at Goangra. For the Border Rivers region it is not possible to completely disentangle the effects of unregulated and regulated recovery on end of system flows, however it is reasonable to assume that the influence of regulated releases will largely occur within the 30 day downstream demand window. The water released by the model to meet the downstream demand is denoted in Figure 67 by the dotted line. It can be seen that the demanded flow pattern is not met exactly due to uncertainty surrounding inflows, losses and travel times. These uncertainties are an important part of day-to-day operations, and are expressed (albeit with a long-term representation) in the models.

Figure 68 Average additional flow in each catchment during demand windows The change in end of system flows was measured for each of the regulated catchments during each demand window throughout the historical climate sequence. Table 43 shows the average additional end of system flows achieved by each catchment to meet downstream (Barwon- Darling) SFIs. Table 40 details these averages along with the standard deviation.

Table 43 Average additional flow in each catchment during demand windows

Change in end of system flows during the demand window (ML/d) Catchment Standard Average Deviation

Border Rivers 530 362

Gwydir 179 104

Namoi 351 236

Page 157

Hydrologic Modelling for the Northern Basin Review

Macquarie–Castlereagh 699 599

This would suggest that downstream releases are having the greatest effect in the Border Rivers and Macquarie–Castlereagh catchments. However, the Gwydir and the Namoi are the only catchments where the increase in end of systems flows can be directly attributed solely to regulated releases. Border Rivers end of system flows are strongly influenced by the recovery of unregulated water entitlements (a large proportion of water recovery in the Border Rivers) while in the Macquarie-Castlereagh regulated release are only made in the Macquarie River however end of system flows are influenced by interactions with other watercourses such as the unregulated Castlereagh and Bogan Rivers (see Figure 69). This is reflected in the higher magnitude and variation in end of system responses for these catchments.

Figure 69 Stlysied representation of the Macquaire-Castlereagh catchment

Page 158

Hydrologic Modelling for the Northern Basin Review

Figure 70 Additional inflows to the Barwon-Darling during demand windows, from Baseline to Northern Standard Each column in Figure 70 represents an individual Barwon-Darling demand that had passed to the regulated tributaries, the height of the column represents the addition inflows to the Barwon- Darling (from Baseline to Basin Plan) during the demand window from all of the regulated tributaries. It can be seen that change to Barwon-Darling inflows due to regulated releases are highly variable. On average, the four tributaries contributed 1,541 ML/d additional inflow to the Barwon–Darling during the demand period. It is worth noting that these values represent the additional end of system flows due to Basin Plan implementation, the combined existing (Baseline) flows during the same demand window periods is 8,331 ML/d, emphasising again that the coordinated delivery approach is supplementing flow events rather than creating new flows. 9.2.2. Event success analysis Section 5.2.3 (Strategy 1) of this report details the modelling process of coordinated tributary releases to meet flow targets in the Barwon-Darling. The current section focuses on an analysis of how these releases impacted flows in the Barwon-Darling. A particular model scenario was generated to assist in the modelling and analysis of downstream environmental water use. This scenario, known as the counterfactual scenario, represents unregulated catchments at Basin Plan conditions with all other catchments with Baseline settings (see Figure 64). An analysis of the outputs of this scenario compared to the Basin Plan scenario demonstrates the influence of regulated releases on the Barwon-Darling flow regime.

Page 159

Hydrologic Modelling for the Northern Basin Review

Figure 71 Example successfully delivered event at Bourke Figure 71 shows an example event at Bourke for which water was requested through downstream demands. In this case, the aim was to achieve a flow above 10,000 ML/d at Bourke for 14 days (this flow event is associated with fish and in-channel environmental outcomes, as all weirs along the Barwon–Darling River have been over-topped and a significant proportion of snags have been inundated). This example demonstrates how the coordinated tributary releases aim to enhance pre-existing events rather than attempting to generate the entire event through regulated releases. For this particular event, the baseline scenario flow (red line) exceeded the desired threshold of 10,000 ML/d for 10 days, hence the desired SFI duration was not met. Increased flow from the unregulated Qld catchments (counterfactual scenario; grey line) extended the duration of >10,000 ML/d flows to 13 days. Regulated releases from the tributaries (390 GL recovery; green line) provided a further increase to 16 days, thereby meeting the SFI criteria.

Page 160

Hydrologic Modelling for the Northern Basin Review

Figure 72 Example failed event at Bourke, the targeted SFI is 6,000 ML/d at Bourke for 14 days This example demonstrates that the combination of water recovery and additional regulated releases can enhance environmental outcomes in the Barwon–Darling. However there are a number of uncertainties associated with achieving this type of enhancement, notably the ability to predict flow travel times and losses between the storages and the inflow point to the Barwon– Darling River. The counter-example in Figure 72 shows an event where the regulated releases failed to result in the targeted SFI being satisfied. In this case, the aim was to achieve a flow exceeding 6,000 ML/d at Bourke for 14 days. Unregulated recovery (grey line) has enhanced the existing event, and, while regulated releases (green line) have augmented the total volume of the event, the SFI parameters were not met. This emphasises the difficulties (both in the model and in reality) associated with anticipating the correct timing of releases across the Northern Basin.

Page 161

Hydrologic Modelling for the Northern Basin Review

Table 44 Requested regulated events success rates

Downstream In-valley demands demands Border Barwon- Gwydir Namoi Macquarie Rivers Darling Number of requested 10 55 14 29 49 events Number of successful 7 23 12 29 22 events Success % 70% 42% 86% 100% 45%

This can be summarised by examining the overall achievement rate for SFI events targeted through regulated releases. Table 44 shows the rate of success for regulated events at Bourke compared to the rate of success for local releases in the regulated tributaries of the Barwon Darling (see section 5.2 for a detailed description of these SFIs). Of the 49 events augmented through releases in the Barwon–Darling, the desired flow shape was achieved in 22 events (i.e. 45%). This does not suggest that the other 55% events were ‘failures’ — the regulated releases still provided environmental outcomes through the wider river system, but these outcomes have not been explicitly detected through the set of SFIs. The Macquarie targets have the highest success rate due to their volumetric nature; i.e. they are less constraining in the desired shape of the hydrograph. The two lowest success rates are related to the Gwydir in-valley targets and the Barwon-Darling coordinated watering targets, however for differing reasons. The success of Gwydir environmental flow targets is impeded by the 300 ML/d flow constraint at Yarraman Bridge. This makes it difficult to meet some of the volumetrically large SFIs in the Gwydir. The low success rate in the Barwon-Darling reflects the difficulties and uncertainties in coordinated environmental watering (i.e. delivering environmental flow events from multiple storages). The analysis performed at the end of system of each regulated tributary was extended to classify the results into ‘SFI achieved’ and ‘SFI unachieved’ events, these results are displayed below in Figure 73. The purpose of this analysis was to examine the relative importance of regulated releases from each catchment.

Page 162

Hydrologic Modelling for the Northern Basin Review

Figure 73 Average chane in flow during downstream demand windows. The average change in end of system flow during a demand period in the Namoi is significantly more during achieved events compared to those that were unachieved. This indicates that the probability of a flow being achieved at Bourke is correlated with the contribution from the Namoi. A similar result is seen for the Border Rivers and the Macquarie, however the magnitude of difference is less than in the Namoi. In the Gwydir the average change in end of system flows is greater for unachieved events indicating that efficacy of regulated releases in the Gwydir have little significance in terms of meeting Barwon-Darling SFIs.

Page 163

Hydrologic Modelling for the Northern Basin Review

Figure 74 Demand activation porportions This analysis can be extended to correlate the achievement of events with the frequency of downstream releases in each catchment. Table 44 shows that the downstream release strategy was activated 49 times over the 114 year modelling period. However, due to the external accounting mechanism underlying the demand approach, it is not possible for a catchment to always contribute additional water downstream. Under the modelled approach, a catchment would be excluded from making additional downstream contributions if either:  there was no water remaining in the environmental account (i.e. the water had alrady been used for in-valley environmental purposes); or,  end of system flows had already exceeded channel capacity (i.e. sending additional flows downstream would lead to overbank flows). Figure 74 shows the proportion of downstream demands that were activated in each tributary catchment classified by the event achievement at Bourke. This difference between the blue and red bars provides an indication of the relative value of regulated contributions from each catchment towards achieving the desired flow in the Barwon– Darling. The presence of regulated contributions from the Namoi is highly correlated with event achievement. Alternatively (e.g. the Gwydir) there is little difference in demand activation between achieved and unachieved events, implying that the regulated contribution from the Gwydir to events at Bourke has little impact on whether the events are achieved or not. Combining the results from Figure 73 and Figure 74 indicates that regulated releases from the Namoi have the greatest capacity to influence desired flows downstream. At the other end of the scale, Barwon–Darling SFI achievement appears to be independent of the contribution from the Gwydir. The contribution from the Border Rivers is moderately important, however it should be noted that the majority of enhanced flows passing downstream from this catchment are the result of

Page 164

Hydrologic Modelling for the Northern Basin Review unregulated recovery rather than regulated releases (only around 30% of water recovered from the Border Rivers model in this scenario is regulated). The conclusion regarding the Macquarie system is more complex. The analysis indicates that the overall presence of regulated environmental releases from the Macquarie is important for achieving the desired Barwon–Darling flow, however the rate of these releases is less important. This is consistent with the connectivity analysis presented in section 3, which noted that the Macquarie contributions downstream generally have a lower peak flow but a longer duration compared to other catchments, a result of the significant flow attenuating influence of the Macquarie Marshes. For all catchments in the Northern Basin, antecedent conditions are an important factor when considering the option of releasing water for downstream purposes. Flow travel time and conveyance costs are highly dependent on the relative wetness of each catchment, and can be difficult to accurately predict. This is especially true for the Macquarie system, which requires significant pre-wetting of the Macquarie Marshes before downstream connectivity is achieved. Environmental releases following a marshes pre-wetting event would help sustain this downstream connectivity, providing enhanced environmental benefits to both the Macquarie and Barwon–Darling systems. 9.3. Downstream Deliverability Conclusions Coordinated delivery of environmental form regulated tributaries to enhance events in the Barwon-Darling requires improved knowledge and operational capacity across the northern Basin. There are many practical challenges to implementing such a strategy. The achievement rate of 45% (of requested events) in the 390 GL scenario demonstrates that the modelling framework reflects the irreducibility of many of the operational challenges and risks. The downstream watering strategy implemented in the NBR modelling is not a suitable as a direct guide for future river operations. However, independent advice has noted that the modelling framework is capable of reasonably simulating the necessary outcomes, and can therefore be used to explore the relative implications of various future scenarios, including the investigation of SDLs (Bewsher 2016). The modelling has included coordinated delivery as an option because the MBDA believe that many of these challenges can be overcome (at least partially) as flow predictive capacity continues to advance. Also important would be the need to modify existing operational practices to allow for downstream water delivery. These changes would need to recognise that downstream delivery could only occur under a specific set of prevailing resource availability and climatic conditions. They would also need to include an element of environmental flow protection within the tributary catchment, while guarding against impacts to other users in the system. Striking this balance would require significant work and engagement with interested parties. Moving towards to more coordinated approach to Northern Basin operations would therefore require a significant investment. As an alternative, the MDBA also modelled an environmental watering strategy that is more in line with current arrangements, and hence would require far less investment in river operations tools. A comparison of the two options (with the same recovery volume) was presented in section 7.2, indicating that a coordinated approach can deliver improved outcomes in the Barwon–Darling. Coordination would therefore provide a more efficient environmental use of recovered water. Further work is required to assess the costs and

Page 165

Hydrologic Modelling for the Northern Basin Review benefits associated with coordinated watering, but it is estimated that a full realisation in the Northern Basin would requires years, possibly decades, to be achieved.

Page 166

Hydrologic Modelling for the Northern Basin Review

10. References Barma (2012) Independent review of models to assess their representation of the baseline conditions specified in the Basin plan and estimating BDLs, Barma Water Resources Pty Ltd Bewsher D (2006b) Namoi Valley, Independent Audit of Cap Model, Prepared for the Murray- Darling Basin Commission, Bewsher Consulting Pty Ltd, 2006. Bewsher D (2009) Peel Valley IQQM, Independent Audit of Cap Model, Prepared for the Murray-Darling Basin Commission. March 2009. Bewsher D (2016) Review of the Hydrological Modelling Frameworks used to inform Potential Basin Plan Amendments, prepared for the Murray–Darling Basin Authority, 2016 CSIRO (2008) Water availability in the Murray–Darling Basin. A report to the Australian Government from the CSIRO Murray–Darling Basin Sustainable Yields Project, CSIRO, Australia. CSIRO (2011) Young WJ, Bond N, Brookes J, Gawne B and Jones GJ, Science review of the estimation of an environmentally sustainable level of take for the Murray-Darling Basin DLWC (1995) Integrated Quantity-Quality Model (IQQM), Reference Manual. NSW Department of Land & Water Conservation, Report No. TS94.048. DERM (2006a) Warrego, Paroo, Bulloo and Nebine, Resource Operations Plan. Department of Natural Resources & Mines, Queensland, January 2006. DERM (2006b) Paroo River Daily IQQM ROP Scenario Modelling Description, System Hydrology Reports (Water Resource Plan), prepared by Surface Water Assessment Group for Resource Management, Department of Natural Resources & Mines, Queensland, August 2006. DERM (2006c) Warrego River Daily IQQM ROP Scenario Modelling Description, System Hydrology Reports (Water Resource Plan), prepared by Surface Water Assessment Group for Resource Management, Department of Natural Resources & Mines, Queensland, May 2006. DERM (2006d) Moonie River Daily IQQM ROP Scenario Modelling Description, System Hydrology Reports (Water Resource Plan), prepared by Surface Water Assessment Group for Resource Management, Department of Natural Resources & Mines, Queensland, May 2006. DERM (2008a) Cap proposal for the Border Rivers valley, Department of Natural Resources & Mines, Queensland, September 2008. DERM (2008b) Moonie Resource Operations Plan, Department of Natural Resources & Mines, Queensland, Reprinted as in Force on February 2006. DIPNR (2004) Namoi River Valley IQQM Cap Implementation Summary Report. Department of Infrastructure, Planning and Natural Resources, Sydney NSW DNR (2006a) Macquarie River Valley IQQM Cap Implementation Summary Report. New South Wales, Department of Natural Resources, 2006. DNR (2006b) Peel River Valley IQQM Cap Implementation Summary Report, NSW Department of Natural Resources, April 2006. MDBA (2011) The proposed “environmentally sustainable level of take” for surface water of the Murray–Darling Basin: Methods and outcomes, MDBA publication no: 226/11

Page 167

Hydrologic Modelling for the Northern Basin Review

MDBA (2012a) Hydrologic Modelling to Inform the proposed Basin Plan — methods and results, MDBA publication no: 17/12 MDBA (2012b) Assessment of environmental water requirements for the proposed Basin Plan: Gwydir Wetlands, MDBA publication no: 32/12 MDBA (2016a) The Northern Basin Review: Understanding the economic, social and environmental outcomes from water recovery in the northern Basin, MDBA publication no: 39/16 MDBA (2016b) The triple-bottom line framework: A method for assessing the, economic, social and environmental outcomes of sustainable diversion limits for the northern basin MDBA (2016c) Environmental Outcomes of the Northern Basin Review MDBA (2016d) Northern Basin Review - Technical overview of the socioeconomic analysis, MDBA publication no: 40/16 Podger G, Yang A, Brown A, Teng J, Power R and Seaton S (2010a) Proposed River Modelling Methods and Integrated River System Modelling Framework Design for use in Basin Plan Modelling. CSIRO: Water for a Healthy Country National Research Flagship. Podger GM, Barma D, Nea, B, Austin K and Murrihy E (2010b) River System Modelling for the Basin Plan Assessment of fitness for purpose. CSIRO: Water for a Healthy Country National Research Flagship, Canberra, December 2010. Webb, McKeown and Associates (2007) State of the Darling, Interim Hydrology Report to the Murray–Darling Basin Commission, ISBN1 921 257 17 2 Yang A. (2010) The Integrated River System Modelling Framework. A report to the Murray- Darling Basin Authority from the CSIRO Water for a Healthy Country Flagship.

Page 168

Hydrologic Modelling for the Northern Basin Review Appendix A: SFI Results Table A1 lists the SFI results for the nine whole-of-north scenarios described in Section 7, along with those for the without development and baseline scenarios. The ‘HU’ and ‘LU’ targets relate to a high and low uncertainty regarding the achievement of the desired environmental outcomes. Also shown in this table is the ‘score’ of each scenario. This scoring approach, in which each SFI was given a value of between 0 and 4 depending on their progress from baseline towards to the HU target value, was developed to collate SFI results. The method underlying this score can be found in the ecological outcomes report (MDBA 2016c).

Table A1: Flow indicator frequency results and score used for aggregated results — Whole-of-north Scenarios

Flow indicator Frequency Score

320 GL 320 GL 320

278 GL 278 350 Scen GL 390 B) (1089; 3 GL 415 GL 278 GL 350 Scen GL 390 B) (1089; 3 GL 415

320 GL (1112; Scen GL 320 E)(1112; Scen GL 320 E)(1112;

321 GL 321 GL 321

Without Without Dev Without Dev

345 GL 345 GL 345

90 GL (1110; Scen 90 GL H) (1110; Scen 90 GL H) (1110;

Baseline Baseline (845) Baseline (845)

GL

SFI

(1111; Scen G) (1111; Scen G) (1111;

(1113; Scen D) (1113; Scen C) (1114; Scen A) (1108; Scen D) (1113; Scen C) (1114; Scen A) (1108;

(1115; Scen J) (1115; Scen J) (1115;

(1103; Scen I)(1103; Scen I)(1103; ID LU HU (844) (844) Flow event

elopment elopment

target target

2 ML/d for 1 day any time of the year at Weilmoringle on the Culgoa River (refuges) (frequency results CB 1 350 430 247 451 447 445 447 448 445 448 447 447 447 4 0 0 0 0 0 0 0 0 0 0 shown as average number of days of top 10% of dry spells) 2 ML/d for 1 day any time of the year at Narran Park on the Narran River (refuges) (frequency results CB 2 350 470 349 542 550 534 533 539 540 540 539 539 539 4 0 0 0 0 0 0 0 0 0 0 shown as average number of days of top 10% of dry spells) 1,000 ML/d for 7 days any time of the year at Brenda CB 3 on the Culgoa River (small fresh) (frequency results 90 80 98 74 75 75 75 75 75 75 75 75 75 4 0 0 0 0 0 0 0 0 0 0 shown as percent of years with at least one event) 1,700 ML/d for 14 days between Aug and May at Wilby Wilby on the Narran River (large fresh) CB 4 60 40 61 25 31 37 40 40 39 40 39 39 40 4 0 1 2 3 3 3 3 2 2 3 (frequency results shown as percent of years with at least one event) 3,500 ML/d for 14 days between Aug and May at Brenda on the Culgoa River (large fresh) (frequency CB 5 60 40 68 30 42 43 46 44 45 41 46 46 46 4 0 3 3 3 3 3 3 3 3 3 results shown as percent of years with at least one event) 9,200 ML/d for 12 days any time of the year at Brenda on the Culgoa River (riparian zone) CB 6 2 3 1.3 5.6 4.0 4.0 3.7 3.9 3.9 3.6 3.5 3.5 3.4 4 0 1 1 2 1 2 2 2 2 2 (frequency results shown as the average period in years between events)

Page 169

Hydrologic Modelling for the Northern Basin Review

Flow indicator Frequency Score

320 GL 320 GL 320

278 GL 278 350 Scen GL 390 B) (1089; 3 GL 415 GL 278 GL 350 Scen GL 390 B) (1089; 3 GL 415

320 GL (1112; Scen GL 320 E)(1112; Scen GL 320 E)(1112;

321 GL 321 GL 321

Without Without Dev Without Dev

345 GL 345 GL 345

90 GL (1110; Scen 90 GL H) (1110; Scen 90 GL H) (1110;

Baseline Baseline (845) Baseline (845)

GL

SFI

(1111; Scen G) (1111; Scen G) (1111;

(1113; Scen D) (1113; Scen C) (1114; Scen A) (1108; Scen D) (1113; Scen C) (1114; Scen A) (1108;

(1115; Scen J) (1115; Scen J) (1115;

(1103; Scen I)(1103; Scen I)(1103; ID LU HU (844) (844) Flow event

elopment elopment

target target

15,000 ML/d for 10 days any time of the year at Brenda on the Culgoa River (inner floodplain) CB 7 3 4 1.9 7.1 6.3 6.3 6.3 6.3 6.3 6.0 5.4 5.4 5.4 4 0 0 0 0 0 1 1 1 1 1 (frequency results shown as the average period in years between events) 24,500 ML/d for 7 days any time of the year at Brenda on the Culgoa River (middle floodplain) CB 8 6 8 3.5 8.7 7.6 7.6 8.1 7.1 7.6 8.1 7.6 7.6 7.6 4 0 3 3 2 3 3 2 3 3 3 (frequency results shown as the average period in years between events) 38,000 ML/d for 6 days any time of the year at Brenda on the Culgoa River (outer floodplain) CB 9 10 20 9.5 28.5 38.0 38.0 22.8 22.8 28.5 22.8 19.0 19.0 16.3 4 0 0 0 2 2 2 2 3 3 3 (frequency results shown as the average period in years between events) 25 GL inflow over 60 days any time of the year at Wilby Wilby on the Narran River (waterbird breeding NL 1 1 1.3 0.6 1.3 1.2 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 4 3 3 4 4 4 4 4 4 4 4 habitat) (frequency results shown as the average period in years between events) 50 GL inflow over 90 days any time of the year at Wilby Wilby on the Narran River (waterbird breeding NL 2 1.3 1.7 0.8 2.0 1.7 1.5 1.4 1.4 1.5 1.4 1.4 1.4 1.4 4 0 3 4 4 4 4 4 4 4 4 and foraging habitat) (frequency results shown as the average period in years between events) 250 GL inflow over 180 days any time of the year at Wilby Wilby on the Narran River (outer floodplain) NL 3 8 10 5.3 13.8 12.2 11.0 9.1 9.1 11.0 10.0 11.0 11.0 9.9 4 0 1 2 3 3 2 3 2 2 3 (frequency results shown as the average period in years between events) 154 GL inflow over 90 days any time of the year at Wilby Wilby on the Narran River (waterbird breeding) NL 4 4 5 2.6 8.3 7.2 6.7 6.3 6.3 7.2 6.3 6.7 6.7 6.3 4 0 0 1 1 1 0 1 1 1 1 (frequency results shown as the average period in years between events) 6,000 ML/d for 14 days any time of the year at BD Bourke on the Darling River (small fresh) (frequency 90 80 96 66 78 80 80 80 82 82 82 75 82 4 0 2 3 3 3 3 3 3 1 3 01 results shown as percent of years with at least one event) 10,000 ML/d for 14 days between Aug and May at BD Bourke on the Darling River (large fresh – fish 80 60 89 54 59 59 59 58 59 59 59 57 60 4 0 2 2 2 2 2 2 2 1 3 04 movement) (frequency results shown as percent of years with at least one event) Two events of 10,000 ML/d for 20 days between Aug and May at Bourke on the Darling River (large fresh – BD05 35 25 42 20 23 22 23 22 22 23 22 22 22 4 0 1 1 1 1 1 1 1 1 1 fish breeding) (frequency results shown as percent of years with at least one event) BD 30,000 ML/d for 24 days any time of the year at 2 3 1.8 4.1 3.9 3.8 3.9 3.9 3.9 3.8 3.8 3.9 3.8 4 0 0 0 0 0 0 0 0 0 0 08 Bourke on the Darling River (riparian zone)

Page 170

Hydrologic Modelling for the Northern Basin Review

Flow indicator Frequency Score

320 GL 320 GL 320

278 GL 278 350 Scen GL 390 B) (1089; 3 GL 415 GL 278 GL 350 Scen GL 390 B) (1089; 3 GL 415

320 GL (1112; Scen GL 320 E)(1112; Scen GL 320 E)(1112;

321 GL 321 GL 321

Without Without Dev Without Dev

345 GL 345 GL 345

90 GL (1110; Scen 90 GL H) (1110; Scen 90 GL H) (1110;

Baseline Baseline (845) Baseline (845)

GL

SFI

(1111; Scen G) (1111; Scen G) (1111;

(1113; Scen D) (1113; Scen C) (1114; Scen A) (1108; Scen D) (1113; Scen C) (1114; Scen A) (1108;

(1115; Scen J) (1115; Scen J) (1115;

(1103; Scen I)(1103; Scen I)(1103; ID LU HU (844) (844) Flow event

elopment elopment

target target

45,000 ML/d for 22 days any time of the year at BD09 3.5 4 3.4 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 4 0 0 0 0 0 0 0 0 0 0 Bourke on the Darling River (inner floodplain) BD 65,000 ML/d for 24 days any time of the year at 6 8 5.6 8.7 8.7 8.7 8.7 8.7 8.7 8.7 8.7 8.7 8.7 4 0 0 0 0 0 0 0 0 0 0 10 Bourke on the Darling River (middle floodplain) 6,000 ML/d for 20 days between Aug and May at BD Louth on the Darling River (small fresh – long 70 70 91 58 61 61 63 62 63 63 63 63 63 4 0 0 0 1 0 1 1 1 1 1 02 duration) (frequency results shown as percent of years with at least one event) 21,000 ML/d for 20 days between Aug and May at BD Louth on the Darling River (large fresh – long 40 40 54 32 32 32 32 32 32 32 32 32 32 4 0 0 0 0 0 0 0 0 0 0 06 duration) (frequency results shown as percent of years with at least one event) 6,000 ML/d for 8 days any time of the year at Wilcannia on the Darling River (small fresh - short BD03 60 45 77 42 46 46 46 48 49 46 48 49 50 4 0 3 3 3 3 3 3 3 3 3 duration) (frequency results shown as percent of years with at least one event) 20,000 ML/d for 5 days any time of the year at BD Wilcannia on the Darling River (large fresh – short 60 45 70 39 42 45 45 45 45 45 45 44 45 4 0 1 3 3 3 3 3 3 2 3 07 duration) (frequency results shown as percent of years with at least one event) Annual flow volume of 2,350 GL measured when flow BD is above 30,000 ML/d at Wilcannia on the Darling 10 7 11 7 7 7 7 7 7 7 7 7 7 4 3 3 3 3 3 3 3 3 3 3 11 River (outer floodplain) (frequency results shown as percent of years with at least one event) 4,000 ML/d for 5 days between Oct and Dec at Mungindi on the Barwon River (fresh – fish breeding) BR 1 31 23 39 17 18 22 20 20 22 22 22 22 22 4 0 0 2 1 1 2 2 2 2 2 (frequency results shown as percent of years with at least one event) 4,000 ML/d for 5 days between Oct and March at Mungindi on the Barwon River (fresh – fish breeding BR 2 59 44 74 33 35 37 37 37 39 38 38 36 39 4 0 0 1 1 1 1 1 1 0 1 longer season) (frequency results shown as percent of years with at least one event) Two 4,000 ML/d for 11 days events any time of the year at Mungindi on the Barwon River (fresh – BR 3 34 25 42 14 14 17 15 79 16 16 18 18 18 4 0 0 0 0 0 0 0 1 1 1 productivity) (frequency results shown as percent of years with at least one event) 150 ML/d for 45 days between Oct and Jan at Yarraman Bridge on the Gwydir River (baseflow – G1 85 85 38 81 79 79 80 84 79 80 80 77 80 0 0 0 0 0 0 0 0 0 0 0 fish movement and breeding) (frequency results shown as percent of years with at least one event)

Page 171

Hydrologic Modelling for the Northern Basin Review

Flow indicator Frequency Score

320 GL 320 GL 320

278 GL 278 350 Scen GL 390 B) (1089; 3 GL 415 GL 278 GL 350 Scen GL 390 B) (1089; 3 GL 415

320 GL (1112; Scen GL 320 E)(1112; Scen GL 320 E)(1112;

321 GL 321 GL 321

Without Without Dev Without Dev

345 GL 345 GL 345

90 GL (1110; Scen 90 GL H) (1110; Scen 90 GL H) (1110;

Baseline Baseline (845) Baseline (845)

GL

SFI

(1111; Scen G) (1111; Scen G) (1111;

(1113; Scen D) (1113; Scen C) (1114; Scen A) (1108; Scen D) (1113; Scen C) (1114; Scen A) (1108;

(1115; Scen J) (1115; Scen J) (1115;

(1103; Scen I)(1103; Scen I)(1103; ID LU HU (844) (844) Flow event

elopment elopment

target target

1,000 ML/d for 2 days between Oct and Jan at Yarraman Bridge on the Gwydir River (fresh – cue for G2 85 85 89 85 84 84 84 82 84 84 84 85 84 4 0 0 0 0 0 0 0 0 4 0 fish movement and breeding) (frequency results shown as percent of years with at least one event) Average of G1 and G2 used to generate the score 85 85 64 83 82 82 82 75 82 82 82 81 82 0 1 0 0 0 0 0 0 0 0 0 used in the report 45 GL inflow over 60 days between Oct and Mar at Yarraman Bridge on the Gwydir River (wetlands) G3 90 80 67 70 75 75 75 61 75 75 75 74 75 0 0 1 1 1 1 1 1 1 1 1 (frequency results shown as percent of years with at least one event) 60 GL inflow over 60 days between Oct and Mar at Yarraman Bridge on the Gwydir River (wetlands) G4 70 60 57 63 62 62 62 47 61 61 61 62 60 0 3 3 3 3 3 3 3 3 3 3 (frequency results shown as percent of years with at least one event) 80 GL inflow over 60 days between Oct and Mar at Yarraman Bridge on the Gwydir River (wetlands) G5 50 40 50 46 46 46 47 25 47 47 47 46 46 4 4 4 4 4 4 4 4 4 4 4 (frequency results shown as percent of years with at least one event) 150 GL inflow over 60 days between Oct and Mar at Yarraman Bridge on the Gwydir River (wetlands) G6 30 20 29 20 25 25 25 14 25 25 25 25 25 4 3 4 4 4 4 4 4 4 4 4 (frequency results shown as percent of years with at least one event) 250 GL inflow over 60 days between Oct and Mar at Yarraman Bridge on the Gwydir River (wetlands) G7 12 12 14 11 14 14 14 79 14 15 15 14 15 4 0 4 4 4 4 4 4 4 4 4 (frequency results shown as percent of years with at least one event) 5.4 GL inflow over 120 days between Feb and Mar and between Aug and Sep at the Mallowa Creek G8 regulator – targeting 50 ML/d during these periods 91 91 17 83 84 84 84 87 87 86 86 84 85 0 0 0 0 0 1 1 1 1 0 0 (riparian veg) (frequency results shown as percent of years with at least one event) 4.5 GL inflow over 92 days between Nov and Jan at the Mallowa Creek regulator – targeting 50 ML/d G9 during this periods (riparian vegetation) (frequency 50 40 15 1 50 50 50 49 49 50 50 51 49 0 0 4 4 4 4 4 4 4 4 4 results shown as percent of years with at least one event) 500 ML/d for 75 days (events with min duration of 25 days included) any time of the year at Bugilbone on N1 55 41 69 33 40 45 45 45 46 46 46 43 45 4 0 2 3 3 3 3 3 3 3 3 the Namoi River (baseflow) (frequency results shown as percent of years with at least one event)

Page 172

Hydrologic Modelling for the Northern Basin Review

Flow indicator Frequency Score

320 GL 320 GL 320

278 GL 278 350 Scen GL 390 B) (1089; 3 GL 415 GL 278 GL 350 Scen GL 390 B) (1089; 3 GL 415

320 GL (1112; Scen GL 320 E)(1112; Scen GL 320 E)(1112;

321 GL 321 GL 321

Without Without Dev Without Dev

345 GL 345 GL 345

90 GL (1110; Scen 90 GL H) (1110; Scen 90 GL H) (1110;

Baseline Baseline (845) Baseline (845)

GL

SFI

(1111; Scen G) (1111; Scen G) (1111;

(1113; Scen D) (1113; Scen C) (1114; Scen A) (1108; Scen D) (1113; Scen C) (1114; Scen A) (1108;

(1115; Scen J) (1115; Scen J) (1115;

(1103; Scen I)(1103; Scen I)(1103; ID LU HU (844) (844) Flow event

elopment elopment

target target

1,800 ML/d for 60 days (events with min duration of 6 days included) any time of the year at Bugilbone on N2 39 29 49 30 32 32 32 32 32 32 32 31 32 4 3 3 3 3 3 3 3 3 3 3 the Namoi River (freshes) (frequency results shown as percent of years with at least one event) 4,000 ML/d for 45 days (events with min duration of 7 days included) any time of the year at Bugilbone on N3 25 22 32 16 19 22 22 22 22 22 22 23 22 4 0 1 3 3 3 3 3 3 3 3 the Namoi River (riparian veg) (frequency results shown as percent of years with at least one event) 100 GL inflow over 5 months between Jun and Apr at Marebone Break on the Macquarie River River M1 85 80 91 80 85 85 84 82 85 85 85 85 86 4 3 4 4 4 3 4 4 4 4 4 (wetlands) (frequency results shown as percent of years with at least one event) 250 GL inflow over 5 months between June and Apr at Marebone Break on the Macquarie River M2 50 40 66 35 48 48 51 46 46 48 48 49 50 4 0 4 4 4 4 4 4 4 4 4 (wetlands) (frequency results shown as percent of years with at least one event) 400 GL inflow over 7 months between June and Apr at Marebone Break on the Macquarie River M3 40 30 48 27 37 37 38 34 37 37 37 40 39 4 0 4 4 4 3 4 4 4 4 4 (wetlands) (frequency results shown as percent of years with at least one event) 700 GL inflow over 8 months between Jun and May at Marebone Break on the Macquarie River M4 17 17 18 17 18 18 18 18 18 18 18 18 18 4 4 4 4 4 4 4 4 4 4 4 (wetlands) (frequency results shown as percent of years with at least one event)

Page 173

Hydrologic Modelling for the Northern Basin Review Table A2: Maximum dry periods between events over the 114-year modelling period — Whole-of-north Scenarios

Maximum Dry Period (years)

SFI Without 278 GL 320 GL 320 GL 321 GL 345 GL 350 GL 390 GL 390 GL 415 GL Flow Event Baseline ID Development (1113; Scen (1112; Scen (1111; Scen (1115; Scen (1103; Scen (1114; Scen (1089; Scen (1110; Scen (1108; Scen (845) (844) D) E) G) J) I) C) B) H) A)

2 ML/d for 1 day any time of the year at Weilmoringle CB 1 on the Culgoa River (refuges) 397 days 712 days 685 days 685 days 685 days 685 days 684 days 685 days 685 days 685 days 685 days

2 ML/d for 1 day any time of the year at Narran Park CB 2 on the Narran River (refuges) 624 days 866 days 862 days 781 days 779 days 854 days 738 days 854 days 854 days 854 days 858 days

1,000 ML/d for 7 days any time of the year at Brenda CB 3 on the Culgoa River (small fresh) 1.7 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5

1,700 ML/d for 14 days between Aug and May at CB 4 Wilby Wilby on the Narran River (large fresh) 5.7 12.7 10.1 10.1 8.0 8.0 8.0 8.0 8.0 8.0 8.0

3,500 ML/d for 14 days between Aug and May at CB 5 Brenda on the Culgoa River (large fresh) 4.4 10.8 10.1 8.0 7.6 10.1 8.0 10.1 7.7 7.7 7.7

9,200 ML/d for 12 days any time of the year at Brenda CB 6 on the Culgoa River (riparian zone) 5.3 28.5 28.5 28.5 19.2 19.2 28.5 19.2 19.2 19.2 19.2

15,000 ML/d for 10 days any time of the year at CB 7 Brenda on the Culgoa River (inner floodplain) 8.9 55.1 55.1 55.1 55.1 55.1 55.1 55.1 28.5 28.5 28.5

24,500 ML/d for 7 days any time of the year at Brenda CB 8 on the Culgoa River (middle floodplain) 10.9 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1

38,000 ML/d for 6 days any time of the year at Brenda CB 9 on the Culgoa River (outer floodplain) 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1

25 GL inflow over 60 days any time of the year at NL 1 Wilby Wilby on the Narran River (waterbird breeding 3.1 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 habitat) 50 GL inflow over 90 days any time of the year at NL 2 Wilby Wilby on the Narran River (waterbird breeding 3.2 7.9 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 and foraging habitat)

250 GL inflow over 180 days any time of the year at NL 3 Wilby Wilby on the Narran River (outer floodplain) 29.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7

154 GL inflow over 90 days any time of the year at NL 4 Wilby Wilby on the Narran River (waterbird breeding) 10.5 29.8 29.8 19.3 19.3 19.3 19.3 19.3 19.3 19.3 19.3

Page 174

Hydrologic Modelling for the Northern Basin Review Table A3: Flow indicator frequency results — Condamine–Balonne Scenarios

Frequency Entitlement Type SFI ID Flow Event Initial SDL Scenarios Spatial Recovery Sensitivity Horizontal Slicing LU HU Sensitivity Target Target 980 1023 1032 1022 1009 1010 1037 1040 1046 1047 1048 1041 1043 1044 2 ML/d for 1 day any time of the year at Weilmoringle on the Culgoa CB 1 River (refuges) (frequency results shown as average number of days 350 430 449 448 448 448 448 447 448 445 422 422 422 448 448 448 of top 10% of dry spells) 2 ML/d for 1 day any time of the year at Narran Park on the Narran CB 2 River (refuges) (frequency results shown as average number of days 350 470 541 548 548 548 540 539 548 534 553 553 553 548 549 548 of top 10% of dry spells) 1,000 ML/d for 7 days any time of the year at Brenda on the Culgoa CB 3 River (small fresh) (frequency results shown as percent of years with 90 80 74 75 75 75 75 75 75 75 75 75 75 75 75 75 at least one event) 1,700 ML/d for 14 days between Aug and May at Wilby Wilby on the CB 4 Narran River (large fresh) (frequency results shown as percent of 60 40 31 38 39 40 39 35 39 39 39 40 39 39 38 39 years with at least one event) 3,500 ML/d for 14 days between Aug and May at Brenda on the CB 5 Culgoa River (large fresh) (frequency results shown as percent of 60 40 38 43 44 45 45 46 44 47 43 44 44 44 42 44 years with at least one event) 9,200 ML/d for 12 days any time of the year at Brenda on the Culgoa CB 6 River (riparian zone) (frequency results shown as the average period 2 3 4.5 3.6 3.6 3.5 3.5 3.4 3.5 3.7 3.9 3.9 3.9 4.2 4.2 4.2 in years between events) 15,000 ML/d for 10 days any time of the year at Brenda on the Culgoa CB 7 River (inner floodplain) (frequency results shown as the average 3 4 6.7 6.0 6.0 5.7 5.7 5.7 5.7 5.7 5.4 5.7 5.4 6.7 6.7 6.7 period in years between events) 24,500 ML/d for 7 days any time of the year at Brenda on the Culgoa CB 8 River (middle floodplain) (frequency results shown as the average 6 8 7.6 7.6 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.1 7.6 7.1 7.6 period in years between events) 38,000 ML/d for 6 days any time of the year at Brenda on the Culgoa CB 9 River (outer floodplain) (frequency results shown as the average 10 20 28.5 22.8 19.0 16.3 16.3 16.3 22.8 22.8 16.3 22.8 16.3 28.5 28.5 28.5 period in years between events) 25 GL inflow over 60 days any time of the year at Wilby Wilby on the NL 1 Narran River (waterbird breeding habitat) (frequency results shown as 1 1.3 1.2 1.1 1.1 1.0 1.1 1.1 1.1 1.0 1.1 1.0 1.0 1.1 1.1 1.1 the average period in years between events) 50 GL inflow over 90 days any time of the year at Wilby Wilby on the NL 2 Narran River (waterbird breeding and foraging habitat) (frequency 1.3 1.7 1.7 1.5 1.4 1.4 1.5 1.6 1.4 1.4 1.4 1.4 1.4 1.4 1.5 1.4 results shown as the average period in years between events) 250 GL inflow over 180 days any time of the year at Wilby Wilby on NL 3 the Narran River (outer floodplain) (frequency results shown as the 8 10 11.0 10.0 10.0 7.8 11.0 11.0 10.0 11.0 10.0 8.4 10.0 10.0 11.0 11.0 average period in years between events) 154 GL inflow over 90 days any time of the year at Wilby Wilby on the NL 4 Narran River (waterbird breeding) (frequency results shown as the 4 5 7.2 6.4 5.7 5.1 5.7 6.0 5.7 6.3 5.6 5.7 5.6 6.0 6.0 6.3 average period in years between events)

Page 175

Hydrologic Modelling for the Northern Basin Review Table A4: Maximum dry periods between events over the 114-year modelling period — Condamine–Balonne Scenarios

Maximum Dry Period (years) SFI ID Flow Event Initial SDL Scenarios Spatial Recovery Sensitivity Entitlement Type Sensitivity Horizontal Slicing 980 1023 1032 1022 1009 1010 1037 1040 1046 1047 1048 1041 1043 1044 685 685 685 685 685 685 685 685 685 685 685 685 685 685 CB 1 2 ML/d for 1 day any time of the year at Weilmoringle on the Culgoa River (refuges) days days days days days days days days days days days days days days 863 858 856 853 857 859 855 779 857 852 856 856 859 857 CB 2 2 ML/d for 1 day any time of the year at Narran Park on the Narran River (refuges) days days days days days days days days days days days days days days

1,000 ML/d for 7 days any time of the year at Brenda on the CB 3 Culgoa River (small fresh) 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5 3.5

1,700 ML/d for 14 days between Aug and May at Wilby Wilby CB 4 on the Narran River (large fresh) 10.1 8.0 8.0 8.0 10.1 10.1 8.0 8.0 8.0 8.0 8.0 8.0 10.1 8.0

3,500 ML/d for 14 days between Aug and May at Brenda on the CB 5 Culgoa River (large fresh) 10.1 10.1 10.1 10.1 10.1 7.7 10.1 7.6 8.0 10.1 10.1 10.1 10.1 10.1

9,200 ML/d for 12 days any time of the year at Brenda on the CB 6 Culgoa River (riparian zone) 28.5 19.2 19.2 19.2 19.2 19.2 19.2 19.2 19.2 19.2 19.2 28.5 28.5 28.5

15,000 ML/d for 10 days any time of the year at Brenda on the CB 7 Culgoa River (inner floodplain) 55.1 28.5 28.5 28.5 28.5 28.5 28.5 28.5 28.5 28.5 28.5 55.1 55.1 55.1

24,500 ML/d for 7 days any time of the year at Brenda on the CB 8 Culgoa River (middle floodplain) 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1

38,000 ML/d for 6 days any time of the year at Brenda on the CB 9 Culgoa River (outer floodplain) 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1 55.1

25 GL inflow over 60 days any time of the year at Wilby Wilby NL 1 on the Narran River (waterbird breeding habitat) 7.5 7.5 7.5 4.9 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5

50 GL inflow over 90 days any time of the year at Wilby Wilby NL 2 on the Narran River (waterbird breeding and foraging habitat) 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5

250 GL inflow over 180 days any time of the year at Wilby Wilby NL 3 on the Narran River (outer floodplain) 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7 54.7

154 GL inflow over 90 days any time of the year at Wilby Wilby NL 4 on the Narran River (waterbird breeding) 29.8 19.3 19.3 19.3 19.3 19.3 19.3 19.3 19.3 19.3 19.3 19.3 19.3 19.3

Page 176

Hydrologic Modelling for the Northern Basin Review

Appendix B: Baseflow Metric Results Baseflows (or low flows) are an important component of the flow regime which maintain aquatic habitats for fish, plants and invertebrates. They comprise the long-term seasonable flows which provide drought refuge during dry periods, and contribute to nutrient dilution during wet periods or after a flood event. A complete description of the baseflow assessment approach adopted for the Basin Plan is given by MDBA (2012a). In summary, a daily baseflow requirement was defined at each site, where this requirement is based on a combination of the without development flow pattern and the 80th percentile without development flow. Consistent with the Basin Plan development modelling, the Northern Basin Review scenarios included the baseflow requirement as demand series in the four catchments with significant public storages (i.e. Border Rivers, Gwydir, Namoi and Macquarie). As described by MDBA (2012a), the achievement of baseflows were assessed through volumetric shortfalls. These are listed in Table B1 for the whole-of-north scenarios completed for the Northern Basin Review. Also listed are the shortfalls as a percentage of total baseflow requirement in Table B2. The sites at which baseflow demands were included are marked grey in both tables. Note that the shortfall values listed in these tables were calculated at a monthly time step, in contrast to the daily time step used during Basin Plan development. This change was adopted because the baseflow requirement series were originally specified at monthly intervals, and a monthly assessment was therefore considered to be more appropriate given the uncertainty inherent to the adopted baseflow methodology. The baseflow results indicate that the implementation of the Basin Plan will improve baseflow outcomes from baseline conditions in the regulated catchments and downstream in the Barwon–Darling. However, beyond 278 GL, the baseflow results demonstrate little dependence on the recovery volume. Over the long-term, baseflows usually only comprise a relatively small volume of the environmental water use, and the modelling indicates that baseflow requirements can largely be achieved with existing recovery (estimated to be 278 GL as of December 2015). Most of the near-unregulated catchments showed little-to-no baseflow shortfalls under baseline conditions, and this was maintained in all Basin Plan scenarios — the shortfall proportions in the Paroo, Warrego and Moonie were all ≤1%. In contrast, the baseflow shortfalls in the Condamine–Balonne are of greater significance, ranging from 1% to 9%. Most notably, the baseflow shortfalls in the Lower Balonne were measured to be 8% (0.7 GL/y; Narran at Wilby Wilby) and 9% (around 1.4 GL/y; Culgoa at Brenda), and these demonstrated little change as a result of water recovery. This is largely an effect of Beardmore Dam which, although a relatively small public storage, absorbs and regulates the majority of upstream low flows. Part of this flow regime is maintained downstream through environmental, stock & domestic releases from Beardmore Dam, but the baseflow data indicates that shortfalls remain in the lower parts of the system. Further examination of the time series showed that the majority of these shortfalls were occurring during dry years, indicating that this may be a risk for Basin Plan outcomes. The shortfalls in Table B1 for these two sites demonstrate little dependence on recovery volume, hence it is likely that baseflow improvement will require an alternative mechanism, such

Page 177

Hydrologic Modelling for the Northern Basin Review as a modification to the pattern of releases from Beardmore Dam during dry years, or a change to translucency rules. There are a number of options available to enhance baseflows through this system, some of which can be achieved under current water sharing arrangements. Further work is required to assess the relative merits of each option.

Page 178

Hydrologic Modelling for the Northern Basin Review Table B1: Baseflow shortfall results under baseline conditions and for the nine whole-of-north scenarios — volumetric shortfall (grey locations indicate that a baseflow demand series was included in the model) Average Shortfall Volume (GL/y) 278 GL 320 GL 320 GL 321 GL 345 GL 350 GL 390 GL 390 GL 415 GL Catchment Site Baseline (1113; (1112; (1111; (1115; (1103; (1114; (1089; (1110; (1108; (845) Scen D) Scen E) Scen G) Scen J) Scen I) Scen C) Scen B) Scen H) Scen A) 424202 – Paroo at Yarronvale 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 424201 – Paroo at Caiwarro 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Paroo 424002 – Paroo at Willara 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 424001 – Paroo at Wanaaring 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 423204 – Warrego at Augathella 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 423203 – Warrego at Wyandra 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Warrego 423202 – Warrego at Cunnamulla 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 423003 – Warrego at Burringun 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 423001 – Warrego at Ford's Bridge 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 422310 – Condamine at Warwick 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 422308 – Condamine at Chinchilla Weir 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Condamine– 422213 – Balonne at Weribone 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Balonne 422015 – Culgoa at Brenda 1.4 1.4 1.4 1.3 1.4 1.4 1.4 1.4 1.4 1.4 422016 – Narran at Wilby Wilby 0.8 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7 Moonie 417001 – Moonie at Gundablouie 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Border 416201 – Macintyre at Goondiwindi 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rivers 416001 – Barwon at Mungindi 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 418029 – Gwydir at Stonybatter 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 418026 – Gwydir at D/S Copeton Dam 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Gwydir 418001 – Gwydir at Pallamalawa 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 418031 – Gwydir at Collymongle 3.2 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 418055 – Mehi at Collarenebri 0.2 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.1 0.0

Page 179

Hydrologic Modelling for the Northern Basin Review

Average Shortfall Volume (GL/y) 278 GL 320 GL 320 GL 321 GL 345 GL 350 GL 390 GL 390 GL 415 GL Catchment Site Baseline (1113; (1112; (1111; (1115; (1103; (1114; (1089; (1110; (1108; (845) Scen D) Scen E) Scen G) Scen J) Scen I) Scen C) Scen B) Scen H) Scen A) 419045 – Peel at Chaffey Dam 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 419015 – Peel at Piallamore 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Namoi 419041 – Namoi at Keepit Dam 4.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 419039 – Namoi at Mollee 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 419026 – Namoi at Goangra 0.3 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 420006 – Castlereagh River at Coonamble 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 420004 – Castlereagh River at Mendooran 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 420001 – Castlereagh River at Gilgandra 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Macquarie– 421007 – Macquarie at Bathurst 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Castlereagh 421040 – Macquarie at D/S Burrendong Dam 2.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 421001 – Macquarie at Dubbo 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 421031 – Macquarie River at Gin Gin 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 421012 – Macquarie River at Carinda 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 422001 – Barwon River at Walgett 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Barwon– 425003 – Darling River at Bourke 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Darling 425008 – Darling River at Wilcannia 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 425012 – Darling River Inflows to Menindee 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

Page 180

Hydrologic Modelling for the Northern Basin Review Table B2: Baseflow shortfall results under baseline conditions and for the nine whole-of-north scenarios — percentage shortfall (grey locations indicate that a baseflow demand series was included in the model) Average Shortfall Volume (Percentage of Total Baseflow Requirement) 278 GL 320 GL 320 GL 321 GL 345 GL 350 GL 390 GL 390 GL 415 GL Catchment Site Baseline (1113; (1112; (1111; (1115; (1103; (1114; (1089; (1110; (1108; (845) Scen D) Scen E) Scen G) Scen J) Scen I) Scen C) Scen B) Scen H) Scen A) 424202 – Paroo at Yarronvale 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 424201 – Paroo at Caiwarro 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Paroo 424002 – Paroo at Willara 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 424001 – Paroo at Wanaaring 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 423204 – Warrego at Augathella 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 423203 – Warrego at Wyandra 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Warrego 423202 – Warrego at Cunnamulla 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 423003 – Warrego at Burringun 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 423001 – Warrego at Ford's Bridge 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% 422310 – Condamine at Warwick 4% 4% 4% 4% 4% 4% 4% 4% 4% 4% 422308 – Condamine at Chinchilla Weir 3% 3% 3% 3% 3% 3% 3% 3% 3% 3% Condamine– 422213 – Balonne at Weribone 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% Balonne 422015 – Culgoa at Brenda 9% 9% 8% 8% 8% 8% 8% 8% 8% 8% 422016 – Narran at Wilby Wilby 8% 7% 7% 7% 7% 7% 7% 7% 7% 7% Moonie 417001 – Moonie at Gundablouie 1% 1% 1% 1% 1% 1% 1% 1% 1% 1% Border 416201 – Macintyre at Goondiwindi 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Rivers 416001 – Barwon at Mungindi 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 418029 – Gwydir at Stonybatter 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 418026 – Gwydir at D/S Copeton Dam 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Gwydir 418001 – Gwydir at Pallamalawa 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 418031 – Gwydir at Collymongle 43% 35% 35% 34% 35% 35% 34% 34% 34% 34% 418055 – Mehi at Collarenebri 1% 0% 0% 0% 0% 0% 0% 0% 0% 0%

Page 181

Hydrologic Modelling for the Northern Basin Review

Average Shortfall Volume (% of total baseflow requirement) 278 GL 320 GL 320 GL 321 GL 345 GL 350 GL 390 GL 390 GL 415 GL Catchment Site Baseline (1113; (1112; (1111; (1115; (1103; (1114; (1089; (1110; (1108; (845) Scen D) Scen E) Scen G) Scen J) Scen I) Scen C) Scen B) Scen H) Scen A) 419045 – Peel at Chaffey Dam 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 419015 – Peel at Piallamore 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Namoi 419041 – Namoi at Keepit Dam 23% 0% 0% 0% 0% 0% 0% 0% 0% 0% 419039 – Namoi at Mollee 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 419026 – Namoi at Goangra 1% 1% 0% 1% 0% 0% 0% 0% 0% 0% 420006 – Castlereagh River at Coonamble 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 420004 – Castlereagh River at Mendooran 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 420001 – Castlereagh River at Gilgandra 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Macquarie– 421007 – Macquarie at Bathurst 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Castlereagh 421040 – Macquarie at D/S Burrendong Dam 8% 0% 0% 0% 0% 0% 0% 0% 0% 0% 421001 – Macquarie at Dubbo 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 421031 – Macquarie River at Gin Gin 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 421012 – Macquarie River at Carinda 2% 1% 1% 1% 1% 1% 1% 1% 1% 1% 422001 – Barwon River at Walgett 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Barwon– 425003 – Darling River at Bourke 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% Darling 425008 – Darling River at Wilcannia 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 425012 – Darling River Inflows to Menindee 0% 0% 0% 0% 0% 0% 0% 0% 0% 0%

Page 182

Hydrologic Modelling for the Northern Basin Review Appendix C: Mass Balance Tables and Graphs

Barwon-Darling

Scenario I Scenario

Scenario J Scenario

Scenario E Scenario B Scenario A Scenario

Scenario D Scenario C Scenario H Scenario

Scenario G Scenario

Water balance (GL/y)

–

–

– – –

– – –

–

(Run #844) (Run #845) (Run

asin Plan Baseline Plan asin

NBR

NBR

NBR NBR NBR

NBR NBR NBR

NBR

B

(320 GL; Run #1112) GL; Run (320 #1111) GL; Run (320 #1115) GL; Run (321 #1103) GL; Run (345 #1114) GL; Run (350 #1089) GL; Run (390 #1110) GL; Run (390 #1108) GL; Run (415

Without Development Without

(278 GL; Run #1113) GL; Run (278

Inflow 4402 2771 2908 2926 2935 2924 2939 2943 2960 2962 2971 Diversions 0 198 162 162 186 157 157 165 166 165 165

Mungindi-Walgett 0 39 32 32 37 31 31 33 33 33 33

Walgett-Bourke 0 101 83 83 95 81 81 84 84 85 85

Bourke-Wilcannia 0 57 47 46 53 45 45 48 48 48 47 Loss* 1310 850 889 892 890 893 896 897 900 902 903 Outflow 3092 1723 1857 1872 1859 1874 1886 1881 1894 1895 1903 * Loss includes unattributed loss

Page 183

Hydrologic Modelling for the Northern Basin Review Annual Barwon-Darling Diversions

Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 350

300

250

200

150

Diversions (GL/y) Diversions 100

50

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 184

Hydrologic Modelling for the Northern Basin Review

Annual Barwon-Darling EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 18,000 16,000 14,000 12,000 10,000 8,000

Flow(GL/y) 6,000 4,000 2,000

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 185

Hydrologic Modelling for the Northern Basin Review

Macquarie- Castlereagh

Baseline Water balance (GL/y)

Scenario D D Scenario E Scenario G Scenario J Scenario I Scenario C Scenario B Scenario H Scenario A Scenario F Scenario

– – – – – – – – – –

Without Development Without #844) (Run Plan Basin #845) (Run NBR #1113) Run GL; (278 NBR #1112) Run GL; (320 NBR #1111) Run PR; GL, (320 NBR #1115) Run GL; (321 NBR #1103) Run GL; (345 NBR #1114) Run GL; (350 NBR #1089) Run GL; (390 NBR #1110) Run GL; (390 NBR #1108) Run GL; (415 NBR #1109) Run GL; (500 Inflow^ 2859 2628 2645 2645 2642 2638 2642 2645 2645 2643 2645 2647 Diversions 0 380 297 297 303 325 305 297 297 297 293 276 Loss* 2099 1671 1745 1745 1739 1719 1737 1745 1745 1745 1748 1762 Outflow 760 577 603 603 600 594 600 603 603 601 604 609 ^ Difference in inflow due to differences in floodplain flow contributions under various scenarios

* Loss includes unattributed loss and change in storage

Page 186

Hydrologic Modelling for the Northern Basin Review Annual Macquarie-Castlereagh Diversions

Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 800 700 600 500 400 300

Diversions (GL/y) Diversions 200 100

0

1925 1895 1900 1905 1910 1915 1920 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 187

Hydrologic Modelling for the Northern Basin Review

Annual Macquarie-Castlereagh EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 8,000 7,000 6,000 5,000 4,000

3,000 Flow(GL/y) 2,000 1,000

0

1895 1990 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1995 2000 2005 Water Year (July to June)

Page 188

Hydrologic Modelling for the Northern Basin Review

#1111)

Namoi-Peel

Water balance (GL/y) #1109) Run

Scenario D D Scenario E Scenario G Scenario J Scenario I Scenario C Scenario B Scenario H Scenario A Scenario F Scenario

– – – – – – – – – –

Without Development Without #844) (Run Baseline Plan Basin #845) (Run NBR #1113) Run GL; (278 NBR #1112) Run GL; (320 NBR Run PR; GL, (320 NBR #1115) Run GL; (321 NBR #1103) Run GL; (345 NBR #1114) Run GL; (350 NBR #1089) Run GL; (390 NBR #1110) Run GL; (390 NBR #1108) Run GL; (415 NBR GL; (500 Inflow 1883 1886^ 1886^ 1886^ 1886^ 1886^ 1886^ 1886^ 1886^ 1886^ 1886^ 1886^ Diversions 0 265 252 245 245 245 241 241 241 241 237 223 Loss* 1055 968 973 976 976 976 976 977 977 977 979 985 Outflow 828 653 661 665 665 665 669 668 668 668 670 678 ^ Includes groundwater gain and Tamworth sewage return

* Loss includes unattributed loss and change in storage

Page 189

Hydrologic Modelling for the Northern Basin Review Annual Namoi Diversions Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 450 400 350 300 250 200

150 Diversions (GL/y) Diversions 100 50

0

1935 1950 1895 1900 1905 1910 1915 1920 1925 1930 1940 1945 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 190

Hydrologic Modelling for the Northern Basin Review

Annual Namoi EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 7,000 6,000 5,000 4,000

3,000 Flow(GL/y) 2,000 1,000

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 191

Hydrologic Modelling for the Northern Basin Review

Gwydir

#1110)

Water balance (GL/y)

Scenario D D Scenario E Scenario G Scenario J Scenario I Scenario C Scenario B Scenario H Scenario A Scenario F Scenario

– – – – – – – – – –

Without Development Without #844) (Run Baseline Plan Basin #845) (Run NBR #1113) Run GL; (278 NBR #1112) Run GL; (320 NBR #1111) Run PR; GL, (320 NBR #1115) Run GL; (321 NBR #1103) Run GL; (345 NBR #1114) Run GL; (350 NBR #1089) Run GL; (390 NBR Run GL; (390 NBR #1108) Run GL; (415 NBR #1109) Run GL; (500 Inflow 996 996 1022 1022 1022 1022 1022 1022 1022 1022 1022 1022 Diversions 0 314 266 266 263 267 267 259 259 259 255 242 Loss* 628 508 555 555 556 554 554 559 559 557 562 571 Outflow 368 174 201 201 203 201 201 204 204 206 205 209 * Loss includes unattributed loss and change in storage

Page 192

Hydrologic Modelling for the Northern Basin Review Annual Gwydir Diversions

Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 800 700 600 500 400 300

Diversions (GL/y) Diversions 200 100

0

1930 1975 1895 1900 1905 1910 1915 1920 1925 1935 1940 1945 1950 1955 1960 1965 1970 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 193

Hydrologic Modelling for the Northern Basin Review Annual Gwydir EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 1,800 1,600 1,400 1,200 1,000 800

Flow(GL/y) 600 400 200

0

1895 1950 2005 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 Water Year (July to June)

Page 194

Hydrologic Modelling for the Northern Basin Review

Border Rivers

#1113)

Water balance (GL/y)

Scenario D D Scenario E Scenario G Scenario J Scenario I Scenario C Scenario B Scenario H Scenario A Scenario F Scenario

– – – – – – – – – –

GL; Run #1110) Run GL;

Without Development Without #844) (Run Baseline Plan Basin #845) (Run NBR Run GL; (278 NBR #1112) Run GL; (320 NBR #1111) Run PR; GL, (320 NBR #1115) Run GL; (321 NBR #1103) Run GL; (345 NBR #1114) Run GL; (350 NBR #1089) Run GL; (390 NBR (390 NBR #1108) Run GL; (415 NBR #1109) Run GL; (500 Inflow 2002 2002 2002 2002 2002 2002 2002 2002 2002 2002 2002 2002 Diversions 0 409 391 380 375 380 367 369 372 370 366 350 Diversions QLD 0 218 202 196 197 196 183 195 195 195 193 185 Diversions 0 191 189 184 178 184 184 174 177 176 173 165 NSW Loss* 1205 1080 1087 1091 1092 1091 1096 1095 1091 1093 1093 1099 Outflow 797 513 524 531 535 531 539 538 539 539 543 553 * Loss includes unattributed loss in the model and change in storage. Loss includes seepage and evaporation from storages and rivers as well as wetland and floodplain habitats.

Page 195

Hydrologic Modelling for the Northern Basin Review Annual Border Rivers Diversions

Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 900 800 700 600 500 400

300 Diversions (GL/y) Diversions 200 100

0

1930 1975 1895 1900 1905 1910 1915 1920 1925 1935 1940 1945 1950 1955 1960 1965 1970 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 196

Hydrologic Modelling for the Northern Basin Review

Annual Border Rivers EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 4,500 4,000 3,500 3,000 2,500 2,000

Flow(GL/y) 1,500 1,000 500

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 197

Hydrologic Modelling for the Northern Basin Review

Moonie

Baseline Water balance (GL/y)

Scenario D D Scenario E Scenario G Scenario J Scenario I Scenario C Scenario B Scenario H Scenario A Scenario F Scenario

– – – – – – – – – –

Without Development Without #844) (Run Plan Basin #845) (Run NBR #1113) Run GL; (278 NBR #1112) Run GL; (320 NBR #1111) Run PR; GL, (320 NBR #1115) Run GL; (321 NBR #1103) Run GL; (345 NBR #1114) Run GL; (350 NBR #1089) Run GL; (390 NBR #1110) Run GL; (390 NBR #1108) Run GL; (415 NBR #1109) Run GL; (500 Inflow 151 151 151 151 151 151 151 151 151 151 151 151 Diversions 0 34 32 32 32 32 29 32 32 32 32 32 Diversions QLD 0 33 31 31 31 31 29 31 31 31 31 31 Diversions 0 1 1 1 1 1 1 1 1 1 1 1 NSW Loss* 55 46 46 46 46 46 47 46 46 46 46 46 Outflow 96 71 73 73 73 73 75 73 73 73 73 73 * Loss includes unattributed loss in the model and change in storage.

Page 198

Hydrologic Modelling for the Northern Basin Review

Annual Moonie Diversions

Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 100 90 80 70 60 50 40

30 Diversions (GL/y) Diversions 20 10

0

1935 1950 1895 1900 1905 1910 1915 1920 1925 1930 1940 1945 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 199

Hydrologic Modelling for the Northern Basin Review Annual Moonie EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 600

500

400

300

Flow(GL/y) 200

100

0

1935 1950 1895 1900 1905 1910 1915 1920 1925 1930 1940 1945 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 200

Hydrologic Modelling for the Northern Basin Review

Warrego

Water balance (GL/y)

Scenario D D Scenario E Scenario G Scenario J Scenario I Scenario C Scenario B Scenario H Scenario A Scenario F Scenario

– – – – – – – – – –

45 GL; Run #1103) Run GL; 45

Without Development Without #844) (Run Baseline Plan Basin #845) (Run NBR #1113) Run GL; (278 NBR #1112) Run GL; (320 NBR #1111) Run PR; GL, (320 NBR #1115) Run GL; (321 NBR (3 NBR #1114) Run GL; (350 NBR #1089) Run GL; (390 NBR #1110) Run GL; (390 NBR #1108) Run GL; (415 NBR #1109) Run GL; (500 Inflow 616 616 616 616 616 616 616 616 616 616 616 616 Diversions 0 52 45 45 45 45 45 45 45 45 45 45 Diversions QLD 0 45 38 38 38 38 38 38 38 38 38 38 Diversions 0 7 7 7 7 7 7 7 7 7 7 7 NSW Loss* 526 488 493 493 493 493 493 493 493 493 493 493 Outflow 90 76 78 78 78 78 78 78 78 78 78 78 * Loss includes unattributed loss in the model and change in storage.

Page 201

Hydrologic Modelling for the Northern Basin Review Annual Warrego Diversions

Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 140

120

100

80

60

Diversions (GL/y) Diversions 40

20

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 202

Hydrologic Modelling for the Northern Basin Review Annual Warrego EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 1,000 900 800 700 600 500 400 Flow(GL/y) 300 200 100

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 203

Hydrologic Modelling for the Northern Basin Review

#1113)

#1108)

Condamine-Balonne

Scenario I Scenario

Scenario J Scenario

Scenario E Scenario B Scenario A Scenario

Scenario D Scenario C Scenario H Scenario

Scenario G Scenario

Water balance (GL/y)

–

#1111)

–

– – –

– – –

–

(Run #844) (Run #845) (Run

NBR

(320 GL, PR; Run Run GL, PR; (320 NBR

NBR NBR NBR

NBR NBR NBR

NBR

Basin Plan Baseline Plan Basin

(320 GL; Run #1112) GL; Run (320 #1115) GL; Run (321 #1103) GL; Run (345 #1114) GL; Run (350 #1089) GL; Run (390 #1110) GL; Run (390 GL; Run (415

Without Development Without

(278 GL; GL; Run (278

Inflow 1707 1700 1701 1701 1701 1701 1701 1701 1701 1701 1701 Diversions (a+b) - 714 648 624 599 599 613 614 572 572 564 Diversions QLD (a) - 713 647 623 598 598 612 613 571 571 563 u/s Beardmore - 286 279 271 268 276 276 276 277 277 277 St. George - 79 79 75 75 75 75 79 74 74 74 St. George to B1 – Water Harvesters - 124 108 106 100 100 109 109 99 99 99 St. George to B1 – OLF & FPH - 18 18 18 18 18 13 18 05 05 05 Narran – Water Harvesters - 24 18 10 04 04 15 04 16 16 10 Narran – OLF & FPH - 08 08 06 06 06 02 09 02 02 02 LBF – Water Harvesters - 117 95 95 90 90 89 106 70 70 68 LBF – OLF & FPH - 57 42 42 37 29 33 12 28 28 28 Diversions NSW (b) - 1 1 1 1 1 1 1 1 1 1 Loss* 1138 744 783 801 818 818 808 809 836 836 841 Outflow 570 242 269 276 284 285 280 278 293 293 296 *Loss includes unattributed loss in the model and change in storage

Page 204

Hydrologic Modelling for the Northern Basin Review

Entitlement Type Initial SDL Scenarios Spatial Recovery Sensitivity Horizontal Slicing Sensitivity Condamine-Balonne Water balance (GL/y)

980

1023 1032 1022 1009 1010 1037 1040 1046 1047 1048 1041 1043 1044

Inflow 1701 1700 1701 1701 1700 1700 1701 1701 1701 1701 1701 1701 1701 1701 Diversions (a+b) 667 614 601 573 573 572 574 575 574 574 574 617 618 622 Diversions QLD (a) 666 613 600 572 572 571 573 574 573 573 573 616 617 621 u/s Beardmore 287 287 287 287 277 266 287 256 287 287 287 287 287 287 St. George 79 79 79 79 79 79 79 79 79 79 79 79 79 79 St. George to B1 – Water 108 115 115 115 114 114 108 73 109 90 104 72 90 72 Harvesters St. George to B1 – OLF & FPH 18 17 17 17 17 17 18 24 0 18 4 18 1 23 Narran – Water Harvesters 21 18 13 0 10 15 10 19 16 3 14 16 16 16 Narran – OLF & FPH 12 0 0 0 1 5 6 10 0 13 2 10 10 10 LBF – Water Harvesters 102 76 77 64 65 66 45 83 82 59 73 99 99 99 LBF – OLF & FPH 39 20 12 10 9 9 20 30 0 25 10 35 35 35 Diversions NSW (b) 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Loss* 771 812 820 838 836 836 834 830 836 831 834 803 805 798 Outflow 263 275 280 290 291 292 293 296 291 296 293 281 278 281 *Loss includes unattributed loss in the model and change in storage

Page 205

Hydrologic Modelling for the Northern Basin Review Annual Condamine-Balonne Diversions

Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 2,500

2,000

1,500

1,000 Diversions (GL/y) Diversions 500

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 206

Hydrologic Modelling for the Northern Basin Review Annual Condamine-Balonne EOS Flow

Without Development Baseline NBR, Scenario B, Run #1089, (390 GL) NBR, Scenario J, Run #1115, (321 GL) 3,500

3,000

2,500

2,000

1,500 Flow(GL/y) 1,000

500

0

1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Water Year (July to June)

Page 207