The need to identify groundwater seepage locations is of great importance for managing both stream water quality and groundwater sourced ecosystems due to their dependency on groundwater-borne nutrients and temperatures. Although several reconnaissance methods using temperature as tracer exist, these are subjected to limitations related to mainly the spatial and temporal resolution and/or mixing of groundwater and surface water leading to dilution of the temperature differences. Further, some methods, for example, thermal imagery and fiber optic distributed temperature sensing, although relative efficient in detecting temperature differences over larger distances, these are labor-intensive and costly. Therefore, there is a need for additional cost-effective methods identifying substantial groundwater seepage locations. We present a method expanding the linear regression of air and stream temperatures by measuring the temperatures in dual-depth; in the stream column and at the streambed-water interface (SWI). By doing so, we apply metrics from linear regression analysis of temperatures between air/stream and air/SWI (linear regression slope, intercept, and coefficient of determination), and the daily water temperature cycle (daily mean temperatures, temperature variance, and the mean diel temperature fluctuation). We show that using metrics from only single-depth stream temperature measurements are insufficient to identify substantial groundwater seepage locations in a head-water stream. Conversely, comparing the metrics from dual-depth temperatures show significant differences; at groundwater seepage locations, temperatures at the SWI merely explain 43–75% of the variation opposed to ⩾ 91% at the corresponding stream column temperatures. In general, at these locations at the SWI, the slopes ( < 0.25) and intercepts ( > 6.5 °C) are substantially lower and higher, respectively, while the mean diel temperature fluctuations ( < 0.98 °C) are decreased compared to remaining locations. The dual-depth approach was applied in a post-glacial fluvial setting, where metrics analyses overall corroborated with field measurements of groundwater fluxes and stream flow accretions. Thus, we propose a method reliably identifying groundwater seepage locations along streambeds in such settings.

Raw project data is available by contacting ctemps@unr.edu