National Water Account 2015

Menu

Region overview

Contextual information

Physical information

Administration

Water rights

Climate and water overview

Water Accounting Statements

Water Assets and Water Liabilities

Changes in Water Assets and Water Liabilities

Water Flows

Notes

Accounting policies

Supporting information

Quantification approach

Reconciliations

Water resources & systems

Water access & use

Future outlook

References

Accountability statement

Murray–Darling Basin: Quantification approaches

Summary of quantification approaches

Table N14 outlines the quantification approaches used to derive the item volumes for the Murray–Darling Basin region. For a more detailed description of the quantification approach, click on the relevant item name in the table.

Table N14 Quantification approaches used to derive item volumes

MDBA = Murray–Darling Basin Authority

Detail of quantification approaches

Water storage product data

Storages

Storage volume at the start and end of the year was calculated using water level data (metres above Australian Height Datum) collected at each storage. Capacity tables established for each storage were used to convert the height measurement to a volume.

The volume of individual storages was aggregated to present the total volume for the item as detailed in the Surface water note. The uncertainty range for these volumes is +/–5%.

The assumptions made were as follows:

• Storage–volume curves represent specifically surveyed parts of the storage and may not reflect the storage–volume relationship across the entire storage.
• Storages are subject to sedimentation and other physical changes over time that in turn affect the accuracy of the storage–volume curves.

Lake capacity data

Lakes and wetlands

The volume of water in lakes and wetlands is based on both measured and estimated data. The volume of water in Lake Burley Griffin at the start and end of the year was calculated using water level data (metres above Australian Height Datum) collected at the lake. Rating tables established for the lake were used to convert the height measurement to a volume.

The volumes of water in Lake Ginninderra and Lake Tuggeranong were estimated based on the known capacities of the lakes, that is, the lakes were assumed to be full at 30 June 2015.

The assumptions made were as follows:

• Water levels in Lake Ginninderra and Lake Tuggeranong are generally managed within 200 mm of full supply level throughout the year. Therefore, the estimated storage volumes of these lakes are considered to be only slightly overestimated.
• The capacity of Lake Burley Griffin is based on survey data collected at the time of construction and fill in 1964.
  
  

AWRA-R model

Regulated and unregulated rivers

The volume of water in the main river channels was modelled using the AWRA-R model, v5.0 (Dutta et al. 2015). In the AWRA-R, a water balance approach is used to calculate the volume of water in a river reach. To the initial volume of water in a reach at the start of a time step, inflow at the upstream nodes, contributing catchment runoff, reservoir contribution, irrigation return, rainfall on the river reach and the return from the floodplain are added and the irrigation diversion, urban diversion (if any), evaporation from the river reach, anabranch flow, overbank flooding, loss to groundwater and the outflow at the downstream node are subtracted to obtain the volume of water in the river reach at the end of the time step. The volume of water in a reach is assumed to be zero at the beginning of the calculation.

The limitations associated with this approach are:

• The modelled ungauged runoff from the AWRA-L was given as input to the AWRA-R model. Therefore, any modelling error in AWRA-L will have some impact on AWRA-R outputs.
• In the AWRA-R model, all the catchment physical processes were not able to be considered due to lack of available data which might give some modelling errors.
  
  

HYDRO database and height-volume relationship

Weirs and locks

River levels were directly measured and converted into volumes using capacity tables for the individual weirs and locks, including locks 6–10 and 15 and Mildura Weir.

Regulated weirs and locks provides a list of all regulated river weirs and locks.

The assumptions and limitations of this approach were:

• The capacity of the lock is taken to be the volume contained in the lock at target storage/full supply level.
• The dead storage associated with the locks is taken to be the volume in storage at the lower level of the operating range.
• This approach allows for comparative measure across years and is preferred to estimating the total volume behind the lock wall, which cannot be accurately measured.
• The accuracy of the capacity tables employed was not evaluated.
• Euston Weir volume excludes the Euston Lakes.

Climate grid data

Precipitation and evaporation

Monthly precipitation grids for the region were produced using daily rainfall data from approximately 6,500 rain gauge stations across Australia and interpolated to a 0.05 degree (~5 km) national grid (Jones et al. 2009).

For the waterbody (storage, lake and weir), the volume of precipitation and transpiration at each waterbody was estimated by multiplying the proportionally weighted average of grid cells that intersected each water feature by the surface area of each waterbody. The average monthly surface area was calculated from daily storage levels and capacity tables where data were available. Where dynamic storage surface area data were not available, the Australian Hydrological Geospatial Fabric (AGHF) surface water feature was used to estimate a static surface area.

For the river sections, the volume of precipitation and transpiration at each river section was estimated by multiplying the average of the grid cells that intersected the contributing catchment for the river section. The daily dynamic surface area was calculated by averaging river width at the inlet and outlet gauging stations, and multiplying by the length of the river section. The river width at a gauging station was obtained based on the flow at the station and the flow-width relationship. The relationship was developed for each gauging station using the cross-section data (Dutta et al. 2015). Where the cross-section data were not available, an approximate estimate of river width was made by examining the neighbouring stations.

The limitations associated with this approach are:

• The precipitation and AWRA-L potential evapotranspiration estimates were subject to approximations associated with interpolating the observation point data to a national grid as detailed in Jones et al. (2009).
• The dynamic storage surface areas calculated from the levels and storage rating tables represent a monthly average and therefore will not capture changes that occur on a shorter temporal scale.
• The use of the static default AHGF surface area is an approximation only. It represents the water features at capacity and therefore is likely to result in an overestimation of precipitation on the water features.
• The accuracy of the river width estimated using the flow-width relationship depends on the quality of the relationship and the type of the function used (Dutta et al. 2015). In AWRA-R, a power law function was used. In addition, the reaches with no cross-section data will have further error due to the use of an approximate estimate of the river width.

Runoff

'Runoff to surface water' was based on streamflow estimates from the AWRA-L version 5.0 model (Viney et al. 2015) outputs. Using gridded climate data for the region (including precipitation, temperature, and solar radiation data), AWRA-L was used to estimate the runoff depth at each grid cell within the region. Only runoff from the landscape was considered; therefore, the surface areas of the major storages were excluded from the analysis.

Runoff from the landscape is divided into two components: runoff into the surface water store (surface water storages, weirs, rivers and drains) and runoff into off-channel water storages. Only runoff into the surface water store was considered here.

The average runoff depth from the landscape into surface water was determined as the weighted mean of the relevant grid-points within the Murray–Darling Basin region. Points were weighted based upon the area they represented within the reporting region to remove edge effects (where the area represented is not wholly within the region) and the effect of changing area represented with changing latitude. Runoff depth was converted to a runoff volume by multiplying runoff depth by the total area of the region (excluding surface water storages, weirs and off-channel storages) and was used as an input to the water balance algorithm.

The assumptions and limitations of this approach were:

• Runoff estimates were compared to historical flows at unimpaired catchments within the region for the 2014–15 year and found to be a reasonable representation of the runoff in this region for this year.
• Runoff estimates were subject to the assumptions of the AWRA-L model (Viney et al. 2015).   
• The estimated runoff corresponds to the runoff expected from an unimpaired catchment. The impairment on runoff from local catchment storages is estimated using a local catchment storage balance model. Where this is applied, the runoff estimates inherit the approximations, assumptions and caveats of the local catchment storage water balance model and the parameters used.

Streamflow data

Outflow

The annual volume of water that flows out to sea from the Murray River is estimated based on data collected at several barrages near the mouth of the river. These data include the flow rates across the barrage and the length of time the barrage is open.

Section 71 Water resource report

Other surface water assets

The volume of surface water stored in Rocky Valley Reservoir, owned by a hydro-electric operator, was provided to the Murray–Darling Basin Authority as part of the s.71 Water Resource Report Victorian input; however, the calculations used to derive this volume were not provided.

The volume of water in Rocky Valley Reservoir is not included in the region's storage volume (see Storages) because no orders can be placed on this storage for delivery of water to the entitlement system. It is only an asset when the hydro-operator physically makes a release.

Allocation remaining

The water management year commences on the date the licence is issued. In most cases, particularly for individual users, the licence anniversary falls outside the standard water year (1 July–30 June). As a result, the water allocation remaining at the end of the 2014–15 year is the unused component of the annual allocation for the licence. The allocation remaining at 30 June 2015 is calculated as shown in Table N15.

Adjustment and forfeiture

For most licences in the region, the portion of water allocation that has not been abstracted at the end of the licence water year is forfeited (i.e. there is no carryover of entitlements). Therefore, forfeiture is calculated as the total annual allocation for each licence minus the allocation abstraction during the licence water year. Individual user entitlements that are terminated during the year are also considered to be forfeitures.

In some cases, such as the Angas Bremer and Marne Saunders groundwater wells in South Australia, carryover is permitted. That is, a portion of the unused annual allocation can be carried over into the next water year. In these cases, forfeiture is calculated as the total annual allocation minus the allocation abstraction and the carryover.

Allocations

Individual user licences are generally issued for periods of between 1 and 10 years, with an annual abstraction amount specified and with annual compliance arrangements in place.

The maximum amount of abstraction for each year is announced annually and is usually a percentage of the licence entitlement. It is also usually based on a review of storage, river, and aquifer levels in the region at the start of the water management year.

More information on these allocations and the associated water access entitlement is given in the Water rights, entitlements, allocations and restrictions note.

Allocated abstraction

The volume of allocated diversions and extractions of surface water and groundwater are for individual users, environmental purposes and the urban water system. They are based on the licensed water year and derived from metered data.

Non-allocated diversion

The non-allocated diversion of surface water for individual users, environmental purposes and the urban water system are based on a combination of metered and estimates. Where metered data are available, the diversion is calculated as the actual diversion during the year. Where metered data are not available, an estimate is made based on historical usage data or modelled data.

Point return: irrigation

The volume of point return from irrigation schemes within the Murrumbidgee and Victorian Murray river systems were derived from metered flow data. Point return data were not available for the other irrigation areas across the region.

The uncertainty estimate for the Broken Creek irrigation scheme is is +/– 40%. The uncertainty for all other schemes is +/– 5%.

Overflow: landscape

Overbank flood spilling is classified as an unregulated event. The volume is applicable to several jurisdictions within the region; however, volumes were only available for Queensland. The volume of water that overflows from river channels onto the landscape is estimated based on a combination of user returns and local knowledge.

A limitation of this approach is that a portion of the reported volume may also include the volume of water harvested from the landscape. It is not possible to distinguish between the volume of water harvested from the landscape and the volume of water that overflows from the river.

The uncertainty estimate for the volume reported is +/– 40%.

River and floodplain losses

A water balance approach was adopted in calculating river and floodplain leakage, evaporation and errors. The calculations were based on hydrological boundaries of river catchments within the region. Total inflows less total outflows and changes in surface water storage for surface water in a reporting unit (a water resource plan area) was considered as the river loss (river and floodplain leakage, evaporation, and errors) for the same unit.

Section 71 Water resource report and water sharing plans

Other groundwater assets

The extractable volume of groundwater (groundwater asset) was estimated as the sum of :

• sustainable diversion limits (SDL) volumes based on information provided by the Murray-Darling Basin Authority
• basic landowner rights or unlicensed stock and domestic rights (where these are not included in the long–term extraction estimates in the water resource plan)
• the volume of supplementary access licence for New South Wales that was available.

This volume may be an underestimate as relevant groundwater data were not available across the entire region.

Extraction: statutory rights

Groundwater extraction data were obtained from the relevant clauses in the jurisdictional groundwater plans and the s.71 Water resource report. The reported volume is likely to be an underestimate as only limited data were available.

Section 71 Water resource report and Snowy hydro database

Claims: inter-region

Snowy Hydro, located outside the region boundary, must release the required annual release (RAR) each Snowy Hydro water management year to the Tumut and Murray rivers. The RAR is set at the commencement of the Snowy Hydro water year (1 May–30 April). The inter-region claim at the end of the 2014–15 year is calculated as shown in Table N16.

Snowy Hydro RAR is based on the calculations in the Snowy Water Licence. The Snowy Hydro RAR remaining at 30 June adjusts the RAR at 1 May by the estimated progressive releases (ML) from 1 May to 30 June. Inputs to the calculation of the RAR include:

• 1,062 GL for the Murray River and 1,026 GL for the Tumut River—the precorporatisation RAR
• the dry inflow sequence volume
• environmental water savings transferred from the Murrumbidgee River and the Murray River to the Snowy River or the montane streams
• recognition of when Snowy Hydro has made a water release in advance of the Snowy Hydro water year
• transfers between the Tumut River and the Murray River
• recognition of water deals between Snowy Hydro and downstream irrigators (both the borrowings and the subsequent paybacks)
• any other RAR adjustment.

Claim increase: inter-region

Inter-region claim increases for the region include:

• Increases in the RAR that Snowy Hydro was required to deliver to the Tumut and the Murray rivers
• a claim by Grampians Wimmera–Mallee Water in the Glenelg catchment.

The total inter-region claim increase during the 2014–15 year is calculated as shown in Table N17.

The limitations associated with this approach are:

• Volumes associated with the Snowy Hydro claim were estimated at the commencement of the water year and subject to revision and confirmation at the end of the year.
• Inflows to Lake Hume for the River Murray are subject to a formula for calculating the respective New South Wales and Victorian shares, which is 50% between New South Wales and Victoria with some adjustments.

Delivery: inter-region agreement

The volumes of water delivered under the above inter-region agreements were based on metered data.

Claim decrease: inter-region

Inter-region claim decreases for the region associated with the Snowy Hydro's RAR was calculated as the sum of the following components:

• repayment of water deals
• reduction in RAR for Snowy River deals (Mowamba)
• Snowy Water licence–reduction agreed between parties for the 2014–15 year
• relation volume reduction (due to irrigatiors entitlement's reaching full allocations and fullness of downstream storages)
• reserved as directed by the NSW Office of Water under clause 13.2 of the Snowy Water licence to facilitate a potential inter-valley transfer
• allocated for the drought account
• call out of relation volume that occurred in previous water year.

Inter-region inflow

The discretionary flow made by Snowy Hydro to Murray and Tumut rivers.

Jurisdictional water sharing plans

Diversion: statutory rights

The volume of diversion under statutory rights was estimated as equal to the total volume of basic landholder rights detailed in the water sharing plan at the commencement of the plan. No information was available for SDLs outside New South Wales.

Other groundwater increases/decreases

Groundwater increases and decreases resulting from changes in long-term extraction estimates and basic land-owner rights were calculated by the Bureau of Meteorology and based on s.71 Water resource report jurisdictional data.

It was assumed that groundwater basic right is included in the permissible consumptive volume provided in relevant water sharing plans if no information is available to decide otherwise.

Discharge: wastewater system

'Treated wastewater discharged to surface water' was metered at wastewater treatment plant outflow points. The estimated uncertainty is +/– 10%.

Groundwater model

Inter-region flow

'Inter-region inflow into the region' was estimated for the unconfined aquifer (Murray Group Limestone and Parilla Sands) and confined aquifer (Renmark Group) that underlie this boundary.

'Inter-region outflow from the region' refers to the lateral outflow from the aquifers into the Murray Limestone and Renmark Group aquifers near the Murray mouth.

The uncertainty in the field-measured data (e.g. groundwater levels, hydraulic conductivity) was not specified and unknown, and hence the impacts of such uncertainty on the calculated groundwater flow were not estimated.

Inter-region coastal flow

Inter-region coastal inflow into the MDB region boundary was only considered to be significant in the area near the mouth of the River Murray in South Australia.

Inter-region coastal outflow from the region refers to the volume of outflow from the aquifers to the Southern Ocean near the River Murray mouth.

Groundwater flow was calculated using a simple geographic information system (GIS) approach based on Darcy's Law. Groundwater levels were interpolated for seasons using the ArcGIS Topo-to-Raster tool from reduced groundwater levels measured at monitoring bores.

Geofabric (Bureau of Meteorology) was used to estimate aquifer thickness. The hydraulic conductivity values were sourced from the Mallee Prescribed Wells Area–Murrayville Water Supply Protection area groundwater model, Department of Water, Land and Biodiversity Conservation, South Australia (Barnett and Osei-bonsu 2006). The transmissivity values were calculated by multiplying the aquifer thickness with the relevant hydraulic conductivity.

Seasonal groundwater flow grids were derived from groundwater level grids, aquifer thickness and hydraulic conductivity using a modification of the ArcGIS Darcy Velocity tool. Groundwater flow across selected flow boundaries was then calculated using a simple GIS analysis and seasonal values were aggregated for the 2014–15 year.

The assumptions and limitations were as follows:

• Regional flow estimations were provided for the Murray Group Limestone Aquifer, which was chosen to represent the unconfined aquifer and the Renmark Group Aquifer. These were considered to be the main aquifer systems that cross the boundary of the Murray–Darling Basin region.
• Inflows and outflows for the MDB region were assumed to occur at or near the coastline only; all the other boundaries were assumed no-flow boundaries mostly representing a groundwater divide.
• Groundwater levels in the unconfined aquifer were assumed to be 0 mAHD (metres above Australian height datum) along the coastline.
• Groundwater flow from the Great Artesian Basin (GAB) to the Murray–Darling Basin and groundwater abstraction from the GAB were not evaluated for the 2015 Account due to lack of data (although this vertical leakage is recognised to be important in some SDL resource units).

Uncertainty information

• The uncertainty estimate was not quantified.
• The uncertainty in the field-measured data (e.g. groundwater levels, hydraulic conductivity) was not specified and unknown, and hence the impacts of such uncertainty on the calculated groundwater flow were not estimated.
• The regional flow estimations were based on the interpolated groundwater level grids produced using a simple GIS analysis. Use of different interpolation methods may impact on the values of the groundwater level grids and hence the estimated regional flow; however, a comparison of this methodology was carried out using a simple groundwater flow model developed on MODFLOW model (United States Geological Survey 2013). The results from the two methodologies indicated a 6–7% difference.
• Groundwater flow was estimated for a simplified boundary constructed from a series of line segments. Groundwater flow across this boundary was calculated using the method described above. The uncertainty surrounding this simplification was not analysed.

Recharge/discharge: landscape

The groundwater recharge and discharge volume to landscape was calculated in selected SDL units across the region. The SDL areas calculated by the Bureau and DPI Water using WAVES and MODFLOW methods respectively are shown in Figure N11.

Figure N11 Sustainable diversion limit areas for modelled groundwater recharge and discharge to landscape

Bureau method

'Groundwater discharge' was estimated along with diffuse groundwater recharge volumes using the water atmosphere vegetation energy and solutes (WAVES) model (Zhang and Dawes 1998; Dawes et al. 1998). In the recharge calculations, depth to water table was considered for all regions where the depth to water table was shallow. A shallow water table was assumed to be where the depth to the water table was  4 m or less below the ground surface. The shallow water table was interpolated using kriging with external drift and the 9" DEM as a physical constraint following the methodology presented in Peterson et al. (2011). Where the water table was not shallow, free drainage conditions were assumed. For SDL resource units with a shallow water table, the model may produce a net discharge from groundwater over the calculation period. The recharge within the region was determined by summing the spatially interpolated positive recharge estimates.

The assumptions and limitations of this approach were:

• The assumptions made in developing the WAVES model as described in Dawes et al. (1998) were all applicable to the recharge estimations carried out for the MDB region.
• The national land use grid (Australian Government Department of Agriculture 2015) was reclassified to three vegetation classes that include annuals, perennials, and trees.  The major vegetation classes modelled were C3 annual pasture, C3 perennial pasture and eucalypt trees with a grass understorey for SDL areas to the south of 31^oS; and C4 annual pasture, C4 perennial pasture and eucalypt trees with a grass understorey for the SDL resource units to the north of 31^oS.
• Annual recharge was estimated using a shallow water table surface estimated by interpolating measured groundwater levels.

The uncertainty in the input parameters and the corresponding impacts on the modelled recharge values were not studied. The uncertainty of the estimated recharge resulting from different recharge interpolation methods was not estimated.

DPI Water method

Recharge volumes were calculated for selected SDL resource units applying New South Wales groundwater models based on MODFLOW (United States Geological Survey 2013) modelling process. Discharge volumes were calculated where the MODFLOW evapotranspiration routines were activated to represent groundwater discharge.

Groundwater recharge is both an input to and an output from a groundwater model. There is no single method for estimating recharge used in the New South Wales groundwater models; however, several models estimate recharge as a percentage of rainfall. The magnitude of recharge (as a percentage of rainfall) can be adjusted during the calibration of a groundwater model so that the observed groundwater levels are reproduced in model outputs as accurately as possible, typically for a period of around 20 years if data are available.

The assumptions and limitations of this approach were:

• Groundwater models make many assumptions and approximations to represent a water balance (United States Geological Survey 2013).
• Several of the New South Wales groundwater models assume estimation of recharge volume as a percentage of rainfall.

Uncertainty of recharge/discharge estimates was not evaluated for the New South Wales groundwater models.

Recharge/discharge: surface water

Groundwater interactions with surface water (discharge to and recharge from) can be represented in MODFLOW models in several ways. Figure N11 shows the DPI Water SDL modelled areas for which volume is calculated in the region. Options that have been used in the New South Wales groundwater models are the MODFLOW river package and the MODFLOW drain package (United States Geological Survey 2013).

Groundwater flow into the river is modelled when groundwater levels are higher than river water levels and water flow is out of the river when river water levels are higher than groundwater levels. MODFLOW also has a subroutine to represent drains. When this is activated and groundwater levels are above the base of the drain, water flow to the drain is estimated and this water volume is removed from the cell of the groundwater model.

For more details about MODFLOW calculations, see documentation at the MODFLOW website (United States Geological Survey 2013).

The assumptions and limitations of this approach were:

• Groundwater models make numerous assumptions and approximations to represent water balance (refer to the MODFLOW website for more details).
• Estimates of water level in rivers that are input to groundwater models are usually taken to be monthly average levels, and the levels would usually have a high level of uncertainty unless a river gauge is located within the groundwater model cell.

The uncertainty estimate was not quantified. It is currently not feasible to estimate the uncertainty of modelled groundwater recharge from surface water from outputs of a MODFLOW groundwater model.

Change in groundwater stored in aquifers

Change in extractable storage is estimated using a simple geographic information system (GIS) approach based on measured groundwater levels and aquifer properties. Firstly, groundwater levels at the start (1 July 2014) and the end (30 June 2015) of the 2014–15 year were estimated. This was achieved by considering all groundwater level measurements between March 2014–October 2014 and March 2015–October 2015 respectively, and using the measurements closest in time to interpolate the start and end levels. The estimated groundwater levels on the start and end dates were then spatially interpolated to grids using kriging with external drift and the 9" Digital Elevation Model (DEM) as an external driver following the methodology presented in Peterson et al. (2011). The change in volume within the sedimentary area was calculated using these interpolated groundwater level surfaces.

These volumes were multiplied by appropriate specific yield values (Commonwealth Scientific and Industrial Research Organisation and Sinclair Knight Merz 2010a and 2010b) to convert the volume to a change in groundwater storage. Finally, change in storage was only considered within a 10-km mask of each groundwater observation bore to ensure an appropriate influence from the change in each bore and the volume was reported for the water table aquifer only. Groundwater storage outside these buffer areas was assumed constant throughout the year given that there is no data available.

The uncertainty in the field-measured data (e.g. groundwater levels, specific yield) was not specified and hence the impacts of such uncertainty on the change in storage were not estimated.

The assumptions and limitations were as follows:

• Change in groundwater storage outside the buffer areas is assumed zero given that no data is available for calculation.
• Change in groundwater storage was not calculated for confined aquifers. Under normal circumstance, the annual change in storage is considered to be negligible for confined aquifers due to their very low storage coefficient, which is much lower than the specific yield of water table aquifers (Freeze and Cherry 1979, Johnson 1967). As long as confined aquifers remain saturated, changes in piezometric levels (i.e. aquifer pressure) usually cause small changes in water volumes stored in the aquifers; the changes are equivalent to the volumetric expansion/contraction of the water and the pore space.
• The specific yield values used in each water table aquifer are presented in the downloadable table.
• The change in storage estimations were calculated from the interpolated groundwater level grids produced using kriging with external drift and the 9" DEM as an external driver. Use of other interpolation methods may impact the values of the groundwater level grids and hence the estimated values for change in groundwater storage.