Identifying groundwater and surface water interaction zones using fiber-optic distributed temperature sensing

Fiber-optic distributed temperature sensing was used to monitoring stream bed temperature patterns to identify groundwater discharge zones of the Upper Sangamon River, as part of the Intensely Managed Landscapes Critical Zone Observatory. This study is being performed to identify the spatial and temporal patterns of groundwater and surface water interactions. A 1-km-long, multimode fiber optic cable was installed on the river bed of the Sangamon River. The cable, connected to an Oryx DTS device, was deployed in a duplexed single-ended configuration with 1 m spatial resolution and 1-hour interval time. The first test for recording temperature measurements for this project was completed over 14 days in late summer 2017. The data collection was calibrated by ice and hot water baths in the first approximately 55 m of the cable. Series temperature-time data of the river bed shows that temperature changed uniformly, and was associated with diurnal atmospheric temperature fluctuations. There was a time-delay effect that between the atmospheric and riverbed temperatures that is controlled by thermal radiation. The downward-trending curve of riverbed temperatures is consistent with a decrease in average air temperature. The curve indicates that, in the anomaly region, temperature in groundwater and surface water interactions was obviously higher than the average temperature of other regions from 6 a.m. to 6 p.m., and much lower than others during the night time. These effects continued throughout the entire test period, and not impacted by atmospheric or hydrologic changes. It is possible that the deviations may represent effects of groundwater discharge from the river bed, or the hyporheic flow out of the bank and streambed, but validation of these hypotheses require additional investigation. To understand the thermal dynamics of groundwater discharge at various spatiotemporal scales, further research will incorporate the contributions from (1) changes in river flow and water depth over time, (2) local streambed seepage using time series thermal data, and (3) stratigraphic and lithologic changes across the river bed. Furthermore, a hydrological-thermal hybrid model, validated by FO-DTS data, may better represent these spatial and temporal variations, so that such interactions could be studied in different hydrologic settings.