Understanding parafluvial exchange and degassing to better quantify groundwater inflows using 222Rn: The King River, southeast Australia

Rn-222 is an important tracer for quantifying groundwater inflows to rivers, especially where groundwater and surface water have similar major ion and stable isotope geochemistry. Uncertainties in the Rn-222 mass balance arise, however, from not accurately estimating the degree of degassing of Rn-222 to the atmosphere and the extent to which interaction within the parafluvial zone provides an additional source of Rn-222. This study estimates both Rn-222 production in the parafluvial zone and degassing along a 75 km stretch of the King River, Australia, in order to more precisely quantify groundwater inflows. The contribution of Rn-222 from the parafluvial zone (F-p) was estimated using Rn-222 emanation rates from near-river sediments and assessment of the residence time of water in the parafluvial zone from Rn-222 activities and Cl concentrations in water the alluvial sediments. Values of F-p range from 40,400 Bq/m/day in the upper King to 8500 Bq/m/day in the lower King, corresponding to differences in the mineralogy and the volume of the parafluvial zone. The gas transfer coefficient (k) was estimated by matching groundwater inflows to observed increases in river discharge during a period of low discharge; k decreases from 25 day(-1) in the upper King to 3 day(-1) in the lower King. These k values are higher than those from most empirical formulations, probably due to the extensive pool and riffle sections that promote degassing. The King River is gaining in its upper and middle sections but several of the lower reaches are losing. Groundwater inflows on a reach scale are as high as 10 m(3)/m/day and cumulative inflows along the 75 km stretch are up to 19,000 m(3)/day. Groundwater inflows increase proportional to total flow reflecting the response of both groundwater and surface water systems to rainfall. Uncertainties in the calculated groundwater inflows are reduced by independently estimating k from the discharge data; however, calculated inflows are up to 40% higher if parafluvial flow were not taken into account. (C) 2014 Elsevier B.V. All rights reserved.