Statistical analysis and interpretation

The only statistic derived from the water quality data was the mean EC, both equally and flow weighted, to be compared to values from aquifer sampling. There is a large difference in the measured total dissolved solids (TDS) of water in the alluvial aquifer and stream, up to a factor of 20, and so discrimination with salinity data is likely to be robust. The two sampling sets for the Gloucester River, at Gloucester (gauge 208020) and Doon Ayre (gauge 208003), have the longest records of the four gauges and the most EC readings. The values in each of the records are consistent and the flow weighted-mean differs from the arithmetic mean EC by less than 2%, so the simpler arithmetic mean stream value was used.

Assuming all sources of a conservative tracer (such as salinity) can be discriminated, the proportion of each source can be determined in the resulting mix. Parsons Brinckerhoff (2013a) provides mean values of water quality parameters, including EC, for a sample of bores in alluvial aquifers in the Gloucester valley. We can assume that all EC is a result of salt derived from baseflow from the alluvial aquifer if (i) salt delivered in rainfall is a very small component, (ii) that salt wash off in runoff is a very small component and (iii) that EC in the alluvial aquifer is greater than in the stream. Under these conditions the ratio of EC of water in the stream to alluvial aquifer gives an upper estimate of the fraction of baseflow in the stream.

Based on EC ratios, the upper estimate for baseflow contribution to the stream for the Gloucester River at Gloucester (gauge 208020) is 3.0% and for the larger subcatchment Gloucester River at Doon Ayre (gauge 208003) is 4.5%. The estimate with far fewer data points for the Avon River at Wenhams Cox is 13.1% and for the Avon River at Gloucester is 13.9%. Using a baseflow separation technique based on recorded daily and monthly stream flows, Parsons Brinckerhoff (2014) suggests that the baseflow for the Avon River at Waukivory Creek is 6% of total flow, and baseflow in the Gloucester River of 29% at Gloucester and 28% at Doon Ayre.

Given that the alluvial aquifer is saline, and about half of the salinity of the underlying rock aquifer is ten times more saline than the stream, it is unlikely that the baseflow component is large. The estimates based solely on daily or monthly flow values take no account of the physical state of the aquifers, their physical properties or their chemical composition. These estimates are uniformly the largest, suggesting baseflow proportions of up to 29%. The physical estimates based on salinity alone provide an upper estimate with a two-component mixing model, and provide uniformly low estimates of baseflow, typically less than 5% where the longest time series of data are available. These provide values that are consistent with diffuse recharge estimates from modelling studies and are representative of the physical observations.

Last updated:
14 June 2018
Thumbnail of the Gloucester subregion

Product Finalisation date