Scaling up point-in-space heat tracing of seepage flux using bed temperatures as a quantitative proxy

It is challenging to quantify reach-scale surface-water-groundwater interactions, while maintaining the fine-scale spatial resolution required in hyporheic studies. One-dimensional heat-transport modeling was used to simulate streambed fluxes at discrete points using time-series temperature records. A predictive relationship was then developed between point-in-time streambed temperature and modeled flux rates. Flux was mapped at high spatial resolution by applying the predictive relationship to mapped streambed temperatures, which allowed for high-resolution quantification of flux by proxy. Inferred patterns of flux are consistent with morphology and yielded a net flux to a 30-m stream reach of 1.0 L s ^-1. Discharge of saline groundwater (5.7 g L ^-1 Cl ^-) allowed for comparison between the temperature proxy method and geochemical variability. Maximum upwelling locations (>35 cm day ^-1) were spatially coincident with areas of high conductance at the bed interface (5-25 mS cm ^-1). Differences between gross flux estimates from heat and geochemical methods are attributed to differences in the spatial extent over which estimates were derived and limited sensitivity of the temperature-as-proxy method. When bed temperatures are near their inherent limits (groundwater and stream-water temperatures) the flux magnitude can be underestimated. Caution must be used when determining gross, reach-scale fluxes from temperature-as-proxy methods when flux rates are outside the sensitivity limits.