H112-0012
Effect of soil pipes on riverbank denitrification during transient hyporheic exchange caused by a peak flow event
Effect of soil pipes on riverbank denitrification during transient hyporheic exchange caused by a peak flow event
Friday, 11 December 2020
Poster
Abstract:
Channel stage fluctuations cause hydraulic gradients in stream and river banks, leading to bidirectional exchange known as bank storage or “lung model” hyporheic exchange. This phenomenon connects aquatic and terrestrial ecosystems, and can inject nitrate from the channel into the hyporheic zone to be consumed by denitrifiers, leading to permanent removal. Soil pipes are common along riverbanks, and undermine efforts to model flow with classical approaches such as Darcy and Richards equations. We used MODFLOW with the Conduit Flow Package (CFP) to simulate lung model exchange across a riverbank with soil pipes. We used MT3D-USGS to simulate nitrate transport from the channel into riparian groundwater and denitrification in riparian zone soils. We varied soil matrix hydraulic conductivity (K), first order reaction constant (k), and soil pipe density, length, height above channel water surface (WSEL), and recorded denitrification caused by a 10-hour peak flow event in the channel. The addition of a single 1.5-meter soil pipe led to a 76% increase in denitrification relative to no soil pipes under basecase conditions (K=10-4 m/s). The addition of five soil pipes increased denitrification 125% and 155% in medium (k=6 per day) and slow (k=0.6 per day) reacting soils, respectively. Linearly extrapolating our MT3D results to longer lengths of stream, a five-soil-pipe-per-meter density led to as much as 30% nitrate removal in a 10-km section of stream. We also found that when soil pipes are above the WSEL, plumes become trapped as water retreats toward the channel, leading to a 54% increase in denitrification in slow reacting soil. Denitrification in coarse soil (K=10-3 m/s) was heavily reaction-limited, but becomes more transport-limited in finer soils. These results are important for understanding reactive transport mechanisms in stream and riverbank hyporheic zones, with implications for watershed nutrient transport and management.

