Streamwater-groundwater interactions and implications for water sharing plans in unregulated catchments : Hunter Valley, eastern Australia

Abstract

A key objective of the Australian National Water Iriitiative of 2004 (clause 23(x)) is that streamwater and groundwater be managed as a single connected system. The National Water Act of 2007 established water sharing plans as a mechanism for ensuring this integrated management of streamwater and groundwater and for driving the sustainable management of Australia's water resources. This thesis investigates streamwater-groundwater interactions in an unconfined alluvial aquifer system in the Widden Brook catchment in the highly productive upper Hunter Valley, eastern Australia. Widden Brook is a right bank tributary of the Goulburn River and contributes up to 17% of flow in the Goulburn. The catchment lithology consists of a basal unit of impermeable Permian carbonaceous shale and coals which is overlain by Triassic Narrabeen sandstone and Tertiary basalts. The aims of this thesis were to reconstruct the hydrological record in the catchment from estimations of the long-term (1913-2007) monthly rainfall, streamflow, baseflow and salt load, reconstructed from records in neighbouring catchments, in order to: (1) recortstruct the long-term hydrological record in the catchment; (2) quantify streamwater-groundwater interactions under changing hydrological conditions; (3) estimate the water balance and the salt balance; and (4) use this data to evaluate the water sharing plan. A record of evapotranspiration was constructed with an established model using spatially interpolated rainfall, evaporation and vegetation coverage in the catchment. Limited stream hydrographs and rating curves were used to estimate long term baseflow and the chloride balance method was used to estimate groundwater recharge. The mean long term (1913-2007) runoff coefficient was 5.4% of rainfall equivalent to 40 mm of specific discharge. Flows however, were highly variable over this period with runoff (specific discharge) of 8.6% (65 mm) and 3.9% (30 mm) for the periods 1950-1979 and 1980-2007 respectively. This decrease in runoff cannot be explained solely by the decrease in rainfall and possible reasons are advanced. The mean estimated annual specific discharge from the reconstructed streamflow record, of 37 mm for the period 1971-2006, was shown to be in good agreement with that predicted using the spatially distributed GROWEST model of 33 mm. Stream salinity in Widden, as measured by electrical conductivity, EC, increases downstream but this is moderated by inputs of freshwater from the tributaries, particularly Blackwater Creek. The relationship between EC and stream flow is used to estimate the salt load discharged by the stream. The mean annual salt load for the period 1913-2007 was 1,813 tonnes/year or a specific salt yield of 3.7 tonnes/km{u00B2} but is highly variable over this period. Estimates of the cyclic salt inputs to the catchment are used to construct a long term salt budget (1913-2007) for the catchment which shows a 29% increase in salt load discharged to streamwater relative to that recharged to the catchment and up to 69% for the flood-dominated period 1950-1979. Estimation of the fraction of the salt load in streamwater at Widden gauge, due to mineral weathering relative to cyclic salt, is in the range 0.17-0.20. There is a distinct relative change in water chemistry downstream. Streamwater chemistry changes from predominantly Na and K ions just below the confluence of Widden Brook and Blackwater Creek, which is associated with kaolinite clay and Na and CI ions in the terrace groundwater, to predominantly Mg and Ca ions in the stream at Emu Creek associated with interstratified illite/smectite exchangeable clays and HC0{u2083}, Mg, Na and CI ions in the groundwater. Strontium isotope ratios are used to show a distinct geochemical signature in groundwater in the older terrace soils which appear disconnected from the stream compared to the younger alluvial floodplain soils which are highly connected to the stream. The hydrochemistry of the groundwater in the terrace in the mid catchment is characterised by higher EC (1500-2200 uS/cm) and oxidising conditions with lower concentrations of soluble iron and higher sulfate concentrations compared to the alluvial floodplain/streamwater which have much lower EC (300-600 ~uS/cm) and mainly reducing conditions, higher concentrations of soluble iron and lower sulfate concentrations. It is shown using geochemistry and water and isotopic tracers that the major sources of salts in the soils, streamwater and groundwater are derived from: (1) evapotranspiration of the shallow groundwater which concentrates salts in the upper soil zone confirmed from CI:Br mass ratios of soil groundwater; (2) mineral weathering of the parent rock material; and (3) ion-exchange reactions in the subsurface clays in the lower catchment. It is inferred that the change in water chemistry in the lower catchment is driven by the strong recharge of streamwater to the aquifer in the losing reach of the catchment which mobilises the Mg and Ca ions in the exchangeable smectite clays in the unconfined groundwater aquifer. The water sharing plan for unregulated catchments in the Hunter Valley is evaluated in light of the above findings. It is found that the information on which the Plan was based is insufficient to ensure sustainability in the system. Suggestions are made for improving the water sharing plan process.