• Change text size
• Skip to content

Supervising Scientist Division

Search:

• About us
• Contact us

• Home
  • Atmosphere
  • Biodiversity
  • Coasts and marine
  • Heritage
  • Human settlements
  • Land
  • Living sustainably
  • Parks and reserves
  • Water
  • EPBC Act

• Supervising Scientist
  • About us
    • Alligator Rivers Region
    • Legislation
    • Structure of SSD
    • Services
    • Working at SSD
  • Environmental monitoring
    • Ranger - Magela chemical data
    • Ranger - Gulungul chemical data
    • Jabiluka - Ngarradj (Swift Creek) chemical data
    • Ranger - Magela biological data
    • Radon monitoring stations
    • Surface Water Chemistry Monitoring
    • Sampling maps
  • Environmental research
  • Supervision and assessment
    • Working with the Northern Territory Government
    • Environmental incidents
    • Uranium mining in the ARR
  • Keeping the community informed
    • Committees and associations
    • Consultation with indigenous communities
    • Displays
  • Publications

Download

• SSR181 - Cross-sectional and scour and fill changes in the Ngarradj catchment between 1998 and 2003 (PDF - 1332 KB)

Executive summary

To measure large-scale bank erosion in the Ngarradj catchment, in which the Jabiluka mine is located, 56 permanently marked channel cross sections were installed in 1998 on the mine site tributaries (Tributaries North and Central) and at the three Environmental Research Institute of the Supervising Scientist (eriss) gauging stations. The sections have been resurveyed annually during the dry season and the results to 2003 are analysed in this report. To measure scour and fill of the sand-bed streams, 30 scour chains were used at some of the above cross sections. They were installed progressively during the 1998 and 1999 dry seasons and have been measured annually during successive dry seasons. The results to 2003 are also analysed in this report. The period between 1998 and 2003 was characterised by average to above average rainfall.

The cross section results indicated that:

• There were minor changes in the Floodout Reach of Tributary North which had developed before the Jabiluka mine headworks and infrastructure had been constructed.
• The main gully on Tributary North was also initiated before the Jabiluka mine was constructed and is currently developing by nickpoint migration at a rate of 1.56 m/a (metres per annum), gully widening and bed degradation.
• The tributary gully on Tributary North was initiated by erosion of a former track between 1984 and 1987, and is currently developing by gully widening and bed aggradation.
• Tributary Central is currently eroding by a combination of active lateral migration, bed degradation and channel widening and these geomorphic processes were initiated before the construction of the Jabiluka mine.
• Minor changes (< 10%) in all hydraulic geometry parameters occurred at all sections at the East Tributary gauge, except the one at the gauging station, which is continuously disturbed by gauging activities.
• Sand storage is currently occurring in the upper Swift Creek gauge reach with all sections exhibiting a decrease in area and in mean and maximum depth.
• Sand storage also dominated at the Swift Creek gauge although it was not as great as at the upper Swift Creek gauge.

The influence of mean bankfull flow velocity, discharge and specific stream power between 1998 and 2003 on percentage change in bankfull channel geometry parameters over the same period in each study reach was assessed by determining product moment correlation coefficients between these two variables. It was found that:

• There were no significant correlations for the Tributary North main gully and for the upper Swift Creek gauge.
• Percentage change in area, width and maximum depth was significantly correlated with mean bankfull flow velocity and specific stream power on the Tributary North tributary gully. The first two were negatively and the last one positively correlated. The reason for the inverse relationship between percentage change in area and width, and the hydraulic parameters is that unstable sites were those with the smallest area and width that were undergoing active enlargement and widening near the primary nickpoint whereas enlarged cross sections further downstream were more stable. The positive correlation between percentage change in maximum depth and mean bankfull flow velocity and specific stream power indicated that the greatest thalweg scour occurred at sites of greatest hydraulic stress.
• There was only one significant correlation (percentage change in area and mean bankfull discharge) on Tributary Central because the greatest channel changes were caused by lateral migration and hence any hydraulic control on channel changes was masked by planform influences.
• While there were many significant correlations at the East Tributary gauge, extreme values at both tails of the data unduly influenced the significance of the relationships.
• At the Swift Creek gauge percentage change in maximum depth was positively related to mean bankfull flow velocity and specific stream power. Thalweg scour was closely related to hydraulic stress.

The survey data were also used to calculate changes in sediment storage for each measurement reach for each wet season between 1998 and 2003 as well as for the complete period. The channel network was a net sediment source due mainly to high rates of channel erosion in Tributary Central. However, the sediment generated by channel erosion on Tributary Central is not supplied to the main channel of Ngarradj but is stored in the channel and fan of Tributary Central and the anabranch of Ngarradj. When the results for Tributary Central are removed from the catchment estimates, sediment storage dominates. Each measurement reach alternated between a sediment source and a sediment store over time although the trends were not synchronous between reaches.

Average annual scour and fill in each measurement reach for each wet season usually overlap with each other, allowing for plus or minus twice the standard error of estimate of the mean. This indicates that mean annual scour and fill are not significantly different between years. Nevertheless, there were substantial variations between wet seasons with the mean annual scour and fill in the measurement reaches varying from -3 ± 24 mm (scour) to 142 ± 41 mm (fill). Net fill is currently occurring in the Ngarradj catchment and hence the bed is a sand storage. However, the current data are not sufficiently reliable to preclude net scour. Furthermore, net scour was recorded in at least one measurement reach for all years, except the first when less data were available.

Continued annual monitoring of the cross sections and scour chains is recommended while the Jabiluka mine is under long-term care and maintenance so as to provide data against which subsequent mine impacts or the success of mine rehabilitation can be determined.