|
|||||
|
|
Towards the representation of groundwater in the Joint UK Land Environment Simulator |
|
| Authors: | Stamatis-Christos Batelis Mostaquimur Rahman Stefan Kollet Ross Woods Rafael Rosolem |
| Institution: | 1. Department of Civil Engineering, University of Bristol, Bristol, UK;2. Institute of Bio- and Geosciences, Agrosphere (IBG-3), Forschungszentrum Jülich, Jülich, Germany
Centre for High-Performance Scientific Computing in Terrestrial Systems, Geoverbund ABC/J, Jülich, Germany;3. Department of Civil Engineering, University of Bristol, Bristol, UK Cabot Institute, University of Bristol, Bristol, UK |
| Abstract: | Groundwater is an important component of the hydrological cycle with significant interactions with soil hydrological processes. Recent studies have demonstrated that incorporating groundwater hydrology in land surface models (LSMs) considerably improves the prediction of the partitioning of water components (e.g., runoff and evapotranspiration) at the land surface. However, the Joint UK Land Environment Simulator (JULES), an LSM developed in the United Kingdom, does not yet have an explicit representation of groundwater. We propose an implementation of a simplified groundwater flow boundary parameterization (JULES-GFB), which replaces the original free drainage assumption in the default model (JULES-FD). We tested the two approaches under a controlled environment for various soil types using two synthetic experiments: (1) single-column and (2) tilted-V catchment, using a three-dimensional (3-D) hydrological model (ParFlow) as a benchmark for JULES’ performance. In addition, we applied our new JULES-GFB model to a regional domain in the UK, where groundwater is the key element for runoff generation. In the single-column infiltration experiment, JULES-GFB showed improved soil moisture dynamics in comparison with JULES-FD, for almost all soil types (except coarse soils) under a variety of initial water table depths. In the tilted-V catchment experiment, JULES-GFB successfully represented the dynamics and the magnitude of saturated and unsaturated storage against the benchmark. The lateral water flow produced by JULES-GFB was about 50% of what was produced by the benchmark, while JULES-FD completely ignores this process. In the regional domain application, the Kling-Gupta efficiency (KGE) for the total runoff simulation showed an average improvement from 0.25 for JULES-FD to 0.75 for JULES-GFB. The mean bias of actual evapotranspiration relative to the Global Land Evaporation Amsterdam Model (GLEAM) product was improved from −0.22 to −0.01 mm day−1. Our new JULES-GFB implementation provides an opportunity to better understand the interactions between the subsurface and land surface processes that are dominated by groundwater hydrology. |
| Keywords: | flow boundary free drainage groundwater land surface model |