Exploring drivers of groundwater recharge at Konza Prairie (Flint Hills region, Kansas, USA) using transfer function noise models

Abstract

Intermittent streams, which lack continuous surface water flow throughout the year, constitute a large proportion of global waterways and influence downstream water quality and quantity. Stream intermittency has been increasing in the Great Plains over the past 40 years, and projected increased duration and severity of drought may cause more frequent or more severe intermittency in the future. Groundwater can be an important control over stream intermittency, and therefore it is important to understand groundwater-surface water interactions related to intermittent streamflow. We used climatological and hydrological data collected from the N04D watershed at the Konza Prairie Biological Station, a tributary to Kings Creek in northeastern Kansas, as inputs for transfer function noise models to assess potential drivers and their individual contribution to variations in groundwater levels in two groundwater monitoring wells, each drawing water from distinct geological units at Konza Prairie. Models for both wells were able to accurately reproduce historical groundwater levels (model Kling-Gupta Efficiency > 0.65). Results from the model for the upper Eiss Limestone Member indicate, in this unit, precipitation may be a joint driver of both stream levels and groundwater recharge. In the deeper Morrill Limestone Member, our models suggest that diffuse recharge and recharge via preferential pathways predominantly drive groundwater recharge to this unit. Our findings may guide future research on intermittent streams in Kansas and inform resource management.