Evaluating the combined value of cosmic ray neutron sensing and water isotopes to track and quantify groundwater recharge and soil moisture dynamics in vadose zone modelling

Estimating areal groundwater recharge (GWR) rates is crucial to assess the sustainability of groundwater resource use. Estimation methods, including physical measurements, water budget approaches, numerical methods and tracer methods each have their strength and limitations. Jointly, monitoring of different hydrological dynamics (e.g. shallow and deep soil moisture, GW levels) and the simulation of the complex process interactions between groundwater, soil, plants and the atmosphere can lead to a more accurate quantitative and scale-relevant estimation.

Influenced by soil hydraulic properties and meteorological conditions, soil moisture plays a crucial role in controlling water flux partitioning into evapotranspiration (ET) and seepage, which can ultimately contribute to groundwater recharge (GWR). For better understanding the soil moisture dynamics, cosmic ray neutron sensing (CRNS) is an increasingly popular method for continuous SM monitoring beyond the point scale over a footprint of (150-300 m radius) over the depth of the root zone (down to 30-50 cm). Soil water isotopes are natural tracers and, for several decades form part of the toolkit for assessing water fluxes in the vadose zone. The general usefulness of both observations has been evaluated separately and successfully in the dedicated modules of the widely used vadose zone model (HYDRUS 5) in previous studies. However, their combined value is yet to be assessed in tracking quantities and timing of GWR.

Therefore, the overall aim of the present study is to evaluate the usefulness of combining field-scale CRNS based soil moisture information with soil water isotopes (δ²H and δ¹⁸O) measurements in HYDRUS 5 to evaluate GWR dynamics in a highly instrumented agricultural hillslope in NE Germany. The study period focuses on two distinct hydrological years, a relatively drier (October 1^st, 2022 – September 30^th, 2023) and a relatively wetter one with considerable snow input in winter (October 1^st, 2023 – September 30^th, 2024).

We parameterize the vadose zone model for two locations on a hillslope with contrasting distances to the GW table. These differences are expected to directly influence GWR travel times and GW contribution to ET. The upslope location has deeper GW table of 4-6 m below surface and the downslope one features shallow GW table, 0.8 – 2.5m, respectively. On the one hand, we employ field-scale SM estimates resulting from a combination of CRNS and adjacent profile SM (down to 100 cm) timeseries at each location, to estimate site-specific transport parameters. Timeseries of groundwater level measurements are additionally used to constrain the bottom boundary of the model. Secondly, we use profiles of bulk soil water isotopes collected along the hillslope on three occasions (May 2023, January 2024 and May 2024) to constrain transport parameters. The calibrated models are then used to track the fate of infiltrated rainfall to estimate GWR travel times and compare dynamics (quantity and timing of GWR) between the dry and wet year.

The insights gained from this modelling exercise will inform future efforts in GWR estimation and drought monitoring networks in NE Germany and evaluate the usefulness of complementing these with dedicated tracer measurements for better understanding hydrological processes driving GWR.