Measuring and modeling water and nutrient flux between a mid-Missouri stream and forested riparian zone in the central U.S

Few studies have investigated spatiotemporal variations of surface water (SW) – groundwater (GW) interactions (including both hydrologic and nutrient) in the central U.S. Therefore, understanding of riparian zone and stream connectivity is limited in that region. Accurate characterizations of SW-GW interactions will improve process based understanding, which is critical for management and outcome predictions of management scenarios. To improve process based understanding of SW-GW interactions, high-frequency water quantity data (stream flow, groundwater flow and precipitation) were collected (5-min intervals) from four stilling wells and two transects of piezometers (n = 6 each) during the 2011 water year along Brushy Creek, located in Boone County, central Missouri. Weekly water quality data (nitrate (NO3 -), total phosphorous (P), potassium (K) and ammonium (NH4 +) were also collected from stream (n = 4) and piezometers (n = 12). Results indicate that Brushy Creek alternates between being a losing and gaining reach, along the study reach (length = 830 m), but is on average a losing stream (-3 x 10-5 m3 s-1 m -1), with a loss of 28 and 7% of total surface flow to groundwater during winter and spring, respectively. Based on established assessment criteria, GW modeling performance with HYDRUS – 1D was deemed 'Very good ' (NS = 0.95, r2 = 0.99, RMSE = 2.38 cm and MD = 1.3 cm) and should therefore be used by land managers with confidence to predict riparian zone water storage and flow. Annual average SW NO3 - was 0.53 mg L-1, while P, K and NH4 + concentrations were 0.13, 3.29 and 0.06 mg L-1, respectively. Nine meters from the stream, annual average concentration for GW NO3 - was 0.01 mg L-1, while total P, K and NH4 + concentrations were 0.03, 1.7 and 0.04 mg L-1, respectively. Results of a hyperbolic model, used to quantify hydrological controls on stream water nutrient concentrations, indicated that NO3- and K exhibited dilution behavior while NH4+ had a concentration effect and P was hydrologically constant. Spatial variations in SW nutrient concentrations varied significantly (p 0.05). Shallow GW modeling with MODFLOW provided numerical approximations of hydrologic and nutrient flux, that are comparable (NS = 0.47, r2 = 0.77, RMSE = 0.61 cm and MD = 0.46 cm) to field observations. Study results indicate that karst geology promotes rapid water movement that can increase dominance of shallow-groundwater geochemical nutrient cycling pathways (e.g. weathering and transport) relative to biochemical nutrient cycling pathways (e.g. plant uptake and N-fixation). Baseline data and results of analysis presented in this dissertation will aid in identification, improvement and validation of management tools that will contribute to advancements in stream - riparian zone best management practices, in particular in karst hydrogeological environments.