Linking regional groundwater models with thermal infrared field surveys to characterize spatial patterns of groundwater discharge at the scale of river networks

Groundwater discharge zones are important hydrologic and ecologic niches. Understanding the spatial distribution of these zones at the watershed scale is key to protecting critical habitat and managing contaminant inputs to streams. Current methods for identifying spatial patterns of groundwater discharge are limited; many common field methods (seepage meters, tracers, etc.) are impractical at the watershed scale and spatial patterns of discharge from watershed-scale groundwater flow models are largely unevaluated. In addition, groundwater models have many possible parameters, and it is unclear which parameters drive modeled spatial patterns of discharge. Despite these challenges, improving our understanding of groundwater discharge patterns at the watershed scale is critical to moving the field forward. We integrated a groundwater flow model and thermal infrared imagery (TIR) surveys to identify spatial patterns of groundwater discharge in a 1570 km² watershed (Farmington River, CT and MA, USA). TIR allowed us to quickly (1-8 km stream length / survey day) identify groundwater discharge areas based on their thermal signature (cold anomalies). We implemented multiple scenarios of a groundwater flow model (MODFLOW-NWT), varying parameters that we predicted drive modeled patterns of groundwater discharge (terrestrial hydraulic conductivity, riverbed conductivity, model resolution). We present results showing the effects of varying model parameters on spatial patterns of groundwater discharge and evaluate those patterns using the TIR. This approach can inform modeling and empirical studies of GW-SW interactions at the watershed scale.