A comparison of fibre-optic distributed temperature sensing to traditional methods of evaluating groundwater inflow to streams

There are several methods for determining the spatial distribution and magnitude of groundwater inputs to streams. We compared the results of conventional methods [dye dilution gauging, acoustic Doppler velocimeter (ADV) differential gauging, and geochemical end-member mixing] to distributed temperature sensing (DTS) using a fibre-optic cable installed along 900 m of Ninemile Creek in Syracuse, New York, USA, during low-flow conditions (discharge of 1·4 m ³ s ^-1). With the exception of differential gauging, all methods identified a focused, contaminated groundwater inflow and produced similar groundwater discharge estimates for that point, with a mean of 66·8 l s ^-1 between all methods although the precision of these estimates varied. ADV discharge measurement accuracy was reduced by non-ideal conditions and failed to identify, much less quantify, the modest groundwater input, which was only 5% of total stream flow. These results indicate ambient tracers, such as heat and geochemical mixing, can yield spatially and quantitatively refined estimates of relatively modest groundwater inflow even in large rivers. DTS heat tracing, in particular, provided the finest spatial characterization of groundwater inflow, and may be more universally applicable than geochemical methods, for which a distinct and consistent groundwater end member may be more difficult to identify.