Characterizing the Hydrologic Impacts of Mountaintop Mining Using Stable Isotopes

Despite mountaintop removal mining (MTM) accounting for the largest land-use change in the Appalachian region of the eastern US, its impact on runoff processes is poorly understood. Several devastating floods have been attributed to MTM activities upstream but there is little quantifiable evidence on how MTM impacts mechanisms of streamflow generation and flooding downstream. MTM involves removing the forest, topsoil, and overlying bedrock to gain access to deeper coal seams. Excess rock is pushed into adjacent valley to create valley fills that completely bury headwater streams that permanently alter ecosystem organization and processes. Isotope hydrology can provide process-based information about the temporal and geographic sources of runoff and rainfall-runoff relationships, but these approaches have not been applied in systems undergoing rapid change and typically not at larger landscape scales. In this study we examine runoff generation using stable isotopes of water from Sycamore Creek (27 km2), an undisturbed forested catchment, and White Oak Creek (11 km2), a MTM-impacted catchment, to quantify for the first time how landscape-scale disturbances impact rainfall-runoff relationship and the processes that govern runoff generation. Both catchments are headwaters of the Clear Fork River watershed (163 km2), an extensively mined and recurrent flood-prone watershed in southern West Virginia, USA. Mountaintop mining in White Oak Creek has disturbed 3 km2 (27% of catchment area) to include 10 valley fills comprising ~0.8 km2 (7%). Stream and rainfall were continuously measured at the outlet of each catchment and water samples were collected using Isco automated water samplers to incrementally characterize isotopic variations in 18O and 2H. Streamflow was separated into event and pre-event water using a two-component hydrograph separation model. The total fraction of event/pre-event water for each event was estimated by linear interpolation between incremental samples of stream and precipitation from the onset of precipitation until stream isotope values returned to pre-event levels. Incremental sampling allows us to estimate the total, peak, and temporal variations of event water contribution during storm events. Our results show that streamflow in White Oak Creek is primarily dominated by event water, whereas pre-event, older water dominates stormflow in the undisturbed Sycamore Creek catchment. We hypothesize that streamflow generation in White Oak Creek is dominated by infiltration-excess overland flow that rapidly delivers event water to the stream, compared to predominantly subsurface flow paths in Sycamore Creek. On-going research using geochemical characterization, end-member mixing analysis, and transit time modeling is aimed at quantifying how MTM impacts the stores, flow paths, and transit times of catchment water.