STABLE ISOTOPES OF WATER SHOW IMPORTANCE OF IMPOUNDED WATER BODIES IN MAINTAINING RIVERINE BASEFLOW

Stable isotopes show that significant flow maintenance is performed by open water bodies during baseflow across scales in New England. Riverine inter-storm baseflow maintains in-stream flow requirements for aquatic habitat and provisioning of human needs and a significant and measurable proportion of flow maintenance is provided by constructed and natural surface water impoundments that form wetlands, ponds, lakes, and reservoirs. Open water evaporation results in a characteristic deviation of the stable isotopic composition of a water body (enrichment of heavy isotopes) that permits partitioning of these sources from groundwater, which tends towards the weighted annual mean isotopic composition of precipitation. We integrated a recent global model of isotopic composition in precipitation, the Regionalized Cluster-based Water Isotopes Prediction (Terzer 2012, IAEA 2013), with regional climate data (the North American Regional Reanalysis) to compare against two datasets of river isotopic composition: 14 New England watersheds from Kendall and Coplen, 2001; and 24 southern New Hampshire subcatchments from Frades (2009) and Zuidema (2012). River water during baseflow was consistently isotopically enriched relative to catchment integrated precipitation composition from RCWIP. The RCWIP performed well against isotopic samples of precipitation from three locations in New England. Variability in catchment enrichment across the region, represented by the isotopic throughflow index (Gibson and Reid 2010), correlates with NLCD open water (r²=0.42), NLCD wetlands (r²=0.25), and total NLCD open water and wetlands (r²=0.60). The stronger correlation against the combined wetland and open-water data, compared to either data source alone, suggests that both land-covers are imparting significantly to catchment non-equilibrium fractionation during catchment drainage. These findings are consistent over a range of scales (on the order of 1 km² to 1000 km²) and suggest that 1) river network scale models should be incorporating the effects of impounded water bodies across scales to adequately define baseflow generation processes for predictive or forecasting studies, and 2) river water isotopes are a potential tool for assessing performance of model baseflow generation processes.