Quantifying Groundwater Flow to a Subtropical Spring-fed River Using Automated 222Rn Measurement

Wednesday, 17 December 2014
Mitra B Khadka and Jonathan B Martin, University of Florida, Gainesville, FL, United States
The magnitude of groundwater discharge to streams can alter stream water chemistry, thereby affecting riverine ecosystems and surface water quality. Point groundwater discharge to streams can be measured using a variety of techniques; however, integrating point and diffuse discharge is difficult over large stream reaches. We applied an automated radon-in-water technique for continuous measurements of 222Rn activities along a 5 km length of the spring-fed Ichetucknee River in north-central Florida. Integration of longitudinal 222Rn distribution, measured on three separate occasions, with groundwater and spring water end members in a mass balance equation allowed temporal and spatial assessment of groundwater flow to the stream. The 222Rn activities indicate groundwater fluxes are higher in the upper reach of the river, which has a narrow flood plain, than in the lower reach, with a wide flood plain. A wide flood plain enhances evapotranspiration, which may cause the observed difference in groundwater seepage. Groundwater flow to the upper reach increases following rain events as diffuse recharge within the catchment increases hydraulic gradients toward the river. Groundwater recharge to the lower reach is smaller and less variable than the upper reach regardless of the river flow. The lower reach can back flood when the Santa Fe River, the receiving stream, floods because of the low gradient of the Ichetucknee River (<2 m/km). Back flooding reduces flow, increases water level and inundates the floodplain, reducing the hydraulic head gradient and groundwater inflow. Based on the 222Rn mass balance, cumulative groundwater inflow is estimated to be 2.5 ± 1 m3/s (±SD) during low flow and 3.2 ± 1.5 m3/s during high flow. The estimated ground water inflows to the Ichetucknee River from the 222Rn mass balance are about twice the estimates of 1.2 m3/s and 1.5 m3/s obtained from dye tracer and ionic chemical tracer methods, respectively. The estimated higher fluxes from 222Rn method than the other techniques could reflect an additional input of 222Rn to the stream, possibly through the hyporheic exchange. The 222Rn mass balance technique appears useful for distinguishing relative amounts of groundwater and hyporheic inflow when combined with the other techniques.