A Validation of Saturated Hydraulic Conductivity Models: Improving Predictive Confidence of Shallow Groundwater-Surface Water Flow Response to Precipitation Events

Abstract:

Saturated hydraulic conductivity (K_sat) estimates are fundamental for characterizing shallow groundwater processes and predicting hydrologic responses to precipitation events. However, confidence in K_sat estimates is often limited by insufficient measurements and incomplete aquifer information. Thus, there is an ongoing need for observed and model validated K_sat values. A scale-nested experimental watershed study was initiated in a representative agricultural catchment of the Chesapeake Bay Watershed in the Northeast USA, to investigate soil physical characteristics, validate simple K_sat models, and quantify shallow groundwater-surface water responses to precipitation events. Soils were characterized using the soil core method and K_sat was quantified with piezometric slug tests. Precipitation data were collected in the catchment every thirty minutes throughout the study period (August 2019-April 2020). Stage/water depth data were collected at eight co-located stream (stilling well) and piezometric locations. Surface soil (i.e., 0-5 cm) dry bulk density and porosity were significantly (p < 0.05) higher and lower, respectively, than deeper soils (i.e., 25-30 and 45-50 cm). Soil characteristics and K_sat differed significantly (p < 0.05) with location. Four out of five tested models showed that spatial variability in farm-scale K_sat estimates was small (CV < 0.5) and validated the use of simple, particle size-based models to accurately predict K_sat. Most hydroclimate variables were significantly different (p < 0.05) between the study period seasons. Shallow groundwater table and stream responses were higher in magnitude throughout the dry season and displayed greater inter-site variability during the wet season. Event-based lag time results showed that the shallow groundwater table and stream water depth responded more slowly (5.5–8.5 hr and 1.0–3.5 hr longer, respectively) to precipitation in the wet season, demonstrating seasonal differences in transit time. Results advance understanding of local water transport processes and demonstrate the practicality of simple, K_sat models to improve predictions of shallow groundwater-surface water response to precipitation events.