Remote assessment of water take

The department routinely estimates water taken during significant flow events, which includes water taken through floodplain harvesting. We do this to improve our:

  • remote sensing capability and methodology
  • floodplain harvesting model calibration
  • understanding of floodplain harvesting and supplementary access water take behaviour.

Our analysis uses remote sensing technology (satellite imagery) to observe the presence of water. In combination with storage curves, we can estimate volumes in large on-farm storages. These are dams mostly used to store floodplain harvesting and unregulated water take. They also store metered licensed take, including supplementary access water.

We compare the volume in storage before and after the event to estimate the total volume of water taken. We then compare the volume taken through a meter to the total take. This allows us to better understand the total volumes of stored water, floodplain harvesting and unregulated take.

For a detailed methodology read our report on the Assessment of take and protection of first flush flows PDF, 1495.35 KB .

While remote sensing is a valuable way of capturing data, it has limits as a means of measuring take. There is inherent uncertainty and scientific error in the methods and analysis which limit its use.

While this method is the best currently available, more accurate, real-time information will be available when floodplain harvesting is measured under our Floodplain Harvesting Measurement Policy 2020 (PDF, 2270.23 KB). This metering, measurement and telemetry will improve how we quantify unregulated and floodplain take and how we plan for and enforce water sharing.

Lessons from this assessment will inform the methods for future assessment, as well as future management of flows for critical needs

Learn more about floodplain harvesting.

Figure 1. Images show the surface area of water in an on-farm storage changing from dry (empty) to maximum surface area (full)

Gwydir 2020 potential floodplain harvesting event


A potential floodplain harvesting event on the Gwydir floodplain began in December 2020.

In summary it is estimated that a total of 54.2 GL of water was taken in this event, including 49.1 GL of licensed water (mostly under supplementary licences) taken through a meter. This means that an estimated 5.1 GL was taken that wasn’t metered, some of which may have been through floodplain harvesting as well as other forms of take as further described below.

On-farm storage volumes

We used data from remote sensing (satellite imagery) and aerial survey, to estimate the volumes of water taken during the event. We based this on changes in the presence, and surface area, of water in large on-farm storages.

Satellite imagery from Sentinal and then Planet (higher resolution) was analysed to determine the change in storage volumes during and after permitted supplementary access. Storages contained water before the event. A total of 54.2 GL of water was stored between mid-December 2020 and mid-January 2021. The volume estimates derived from the two different imagery sources were comparable to within 10%. Table 1 presents the results of the assessment.

Table 1. Total volume estimated for on-farm storages (OFS) using higher resolution imagery.
14-Dec-20 to 12-Jan-21
13-Jan to 27-Jan-21
Number of OFS with water156242178
Number of dry OFS252166230
Total volume (GL)36.190.361.5
Total stored / use (GL)-54.2-28.8
Supplementary water

Supplementary access was announced for the Gwydir Valley from the 17th December 2020 until the 7th January 2021. For more information on each supplementary flow announcement visit WaterNSW.

Table 2 below shows the volume of supplementary water take during permitted access. It also shows any other licensed water taken during the floodplain harvesting event. A total of 49.1 GL of metered licensed water was taken during the event. This volume is comparable to the volume of change estimated by the remote sensing method (54.2 GL). The difference is 5.1 GL, which is within the error margin of our methods. This could be unmetered floodplain harvesting and unregulated take, direct and on-farm rainfall run-off/tailwater capture, stock and domestic take, groundwater take or an artefact of uncertainty in the remote sensing method. Table 2 presents the metered licensed water take, including supplementary access, which ended mid-event (7th January 2021).

Table 2. Metered licensed water take.
Use  typeMid-event use
17-Dec-20 to 7-Jan-21 (GL)
Post-event use
8-Jan-21 to 27-Jan-21 (GL)
Total (GL)
Supplementary (consumptive)
Other licence extractions*
Total consumptive extractions48.80.349.1
Held environmental water licence use1.75.16.8

*Provisional figures based on metering data extracted 18 March 2021. 17th December 2020 supplementary access commenced. 7th January 2021 supplementary access completed. Excludes Local water utility use, groundwater and unregulated use and extractions between Copeton Dam and Pallamallawa.

February 2021 event insights


In late February 2021, the Northern Valleys of the Murray Darling Basin (Gwydir, Namoi, Macquarie, Barwon Darling and Border Rivers) received significant rainfall. We closely monitored this event to analyse flows, river extractions, and take from other sources (e.g. floodplain harvesting).

Employing a mix of remote sensing technologies (i.e. satellite observations) and elevation data, we used and refined the method used in the early 2020 ‘First Flush Flows’ , to replicate the same analysis over a similar geography.

In doing so, analysts estimated both the filling period and volumes of water held in private storages for the nominated ‘event’ duration. Combining these analyses with water account and hydrometric station data gives a consolidated event summary.


Results from remote sensing analytics show an observed total event increase from all forms of take of nearly 427 GL of water put into private storages between late-February and mid-April 2021.

This includes 206 GL that was taken through a meter and with a water licence, and around 221 GL from other forms of take, including floodplain harvesting.

These figures are constrained to only storages that could be observed in cloud free imagery, and to only storages with a known storage curve. The volumes are scaled down to account for inaccuracies in the analysis, making it even more conservative. Therefore, the actual stored volumes may be substantially higher in some catchments.

Consolidated totals for the five valleys are in Table 1 below.

Table 1. Floodplain management area volume change summary.
Northern Valleys Pre-event
21-Feb to 13-Mar-21
18 to 31-Mar-21
2 to 14-Apr-21
Number of storages Wet 677 560 1,002
Dry 485 183 165
No image 7 426 2
No curve 245
Total 1,414
Volume Volume in storage (GL) 182 184.1 609.6
Full supply volume (FSV) (GL) 1,179.4
% of FSV 15% 16% 52%

To complement the remote observations, we assessed metering information for the regulated river water sources within the floodplain management areas (Table 2). We used water orders to apportion the measured usage to the periods assessed for storage volume change.

This shows up to 32% (Macquarie), 65% (Namoi), 35% (Gwydir) and 49% (Border Rivers) may have been supplied by regulated river water source extractions. The additional volume change could be made up of unmetered floodplain harvesting and unregulated take, direct and on-farm rainfall run-off/tailwater capture, stock and domestic take, groundwater take, or an artefact of uncertainty and limitations in the remote sensing method. The largest of these potential sources are considered to be unmetered floodplain harvesting and unmetered unregulated river extraction. Overall, across the Northern Basin, the metered water take volumes for Barwon Darling tributaries account for 43% of the total water take estimated by remote sensing (Table 2).

Table 2. Metered river extraction summary
Floodplain management area River extraction water source1 Consumptive licence use (GL)2 Date from Date to Floodplain management area storage change (GL) % of volume change
Border Rivers NSW Border Rivers Regulated 393 3-Mar-21 2-Apr-21 79 49%
Gwydir Gwydir Regulated 82 21-Feb-21 12-Apr-21 232 35%
Namoi Lower Namoi Regulated 41 3-Mar-21 14-Apr-21 63 65%
Macquarie Macquarie and Cudgegong Regulated 7 13-Mar-21 2-Apr-21 22 32%
Total 16921-Feb-2114-Apr-2139643%
Barwon Darling Barwon Darling unregulated 374 1-Mar-21 5-Apr-21 31 >100%
Total 206 21-Feb-215-Apr-21427N/A

1 Usage assessed downstream of Dumaresq–Macintyre junction (Border Rivers), downstream Pallamallawa (Gwydir), downstream Gunnedah (Namoi), downstream Dubbo (Macquarie)
2 Excludes held environmental water use and river extractions from unregulated river water sources other than the Barwon Darling. Extractions for regulated river water sources include water access licence use for supplementary, general security and high security access licence categories.
3 Excludes NSW allocation extracted in QLD. Excludes QLD allocation extracted in NSW (no information was available at the time of analysis)
4 Estimated volume for the remote sensing analysis period of 1 March 21 to 5 April 21. Provisional Barwon Darling consumptive extractions for the 2020-21 water year at time of analysis was 192 GL. This excludes held environmental water access licence orders. The estimate considers potential flow access under the water sharing plan flow rules for the analysis period, and apportions the total use based on days of access. Variance to the flow access rules for environmental orders are not considered in this estimate.


Substantial flow from the event transitioned through the Barwon Darling river system to the Menindee Lakes. The long travel times for flow in this system need a wider period of observation to illustrate the full outcome of such flow events. For the period 1 February 2021 to 30 June 2021 over 1,500 GL flowed through Walgett, and over 1,100 GL through Wilcannia.

Flow gauging stations in the connecting tributaries show large volumes flowing to the Barwon Darling from the Gwydir (406 GL)5, Border Rivers (391 GL)6, Namoi (156 GL)7, and Macquarie, Castlereagh and Bogan rivers (170 GL)8. Volumetrically, these are likely conservative, particularly for the Gwydir, given the large volumes observed in the Barwon Darling, and the limitations of high flow measurement when flows break over the floodplain in the northern valleys. Significant flows were also transitioning to the Barwon Darling from the Culgoa (150 GL)9 and Moonie rivers (100 GL)10.

Within this period the combined Menindee Lakes storage volume increased by 767 GL and was continuing to rise as further flow transitioned through. Additionally, 109 GL was released from the lakes providing flow through to the Murray River.

Northern Basin flows, 1 February 2021 – 30 June 2021.

5 Includes Gil Gil Creek at Galloway (416052), Mehi River at Bronte (418058), Moomin Creek at Moomin Plains (418070), Thalaba Creek at Belarre. Excludes Gingham Channel, where a further 67 GL was measured flowing to the lower Gwydir floodplain
6 Barwon River at Mungindi
7 Includes Namoi River at Goangra, Pian Creek at Waminda
8 Includes Castlereagh River at Gungalman (420020), Macquarie River at Brewon Bridge (gap filled) (421907), Marra Creek at Billybingbone Bridge (421107), Bogan River at Gongolgon (421023)
9 Culgoa River at downstream Collerina (422006)
10 Moonie River at Gundablouie (417001)

While a substantial volume of water was stored through floodplain harvesting and unregulated take, this has occurred during a wet period when there are large volumes of water on land and in rivers, enough to replenish and re-fill weir and refuge pools, wetlands, and the Menindee Lakes. This analysis underscores the department’s management of water resources, including through the protection of first flushes, by understanding and quantifying take, and ultimately permitting floodplain harvesting only during wet periods when downstream critical human and environmental needs will be fully satisfied.