Identifying the major variables controlling transport of water and analytes from an alluvial aquifer to streams

Reliable information on groundwater (GW) lateral flow characteristics is required for estimation of GW extraction, environmental flow requirements, contaminant loading from GW to surface water bodies (SWB), and aquifer remediation purposes. Lateral flow from a shallow alluvial aquifer was investigated applying parametric and non-parametric statistics to flux-theory based outputs obtained using time series hydraulic head (HH) and analyte concentration data. The emphasis of the investigation was to identify the major variables that control the export of contaminants from GW to SWB. Point measurements from 4 shallow wells (10-12 m deep) installed along a 1.1 km transect perpendicularly crossing a creek were undertaken at 7 - 12 day intervals from January through June (wet season) over 3 years in a wet tropical catchment in north-eastern Australia. The HH during two wet seasons at north upslope varied from 4.84 m to 12.37 m with mean, median, and coefficient of variation (CV) of 8.86 m, 8.73 m, and 17% respectively. At the downslope the corresponding values were 3.59-6.21 m, 4.81 m, 4.58 m, and 13%, respectively. Similar temporal trends were observed at the south upslope and downslopes. Nitrate-N concentrations at the north upslope varied from 23 to 1340 µg L-1 with mean, median, and CV of 691 µg L-1, 609 µg L-1 and 23%, respectively. Similar trends were observed at north downslope and at up- and down-slopes of southern transect. The lateral hydraulic gradient (LHG) from north upslope to downslope varied from 4.12 x 10-3 to 9.92 x 10-3 m m-1 and the corresponding flow velocity (Vx) from 3.63 x 10-3 to 3.48 x 10-2 m d-1. Nitrate-N flux from north upslope to downslope varied from 1.0 x 10-4 to 4.4 x10-3 g m-2 d-1, similar trends were observed for EC and Cl and also along southern transect. These suggest that analyte fluxes followed the LHG indicating conservative transport of the former from upslopes to downslopes. The conservative transport was reconfirmed by significant associations between HH and analyte fluxes; R2 18-70% for EC, 24-52% for Cl, and 52-76% for nitrate. Travel time for 650 m, computed using mean Vx varied from 5.8 to 69 yrs and the variations depended on the values of saturated soil hydraulic conductivity (Ks) used. The results indicate contaminant export extrapolations from point measurements to landscape scales depended on our ability to incorporate spatial and temporal variabilities in Vx and analyte fluxes, reliable information in Ks, and macropore bypass flow. We believe this is one of the few studies that have coupled flux-theory and statistics to identify and assess the major variables that control contaminant export from GW to SWB