Extreme Resolution Ecohydrologic Modeling for Understanding Micro-topographic Controls

Abstract:

The advent of extremely high-resolution data such as lidar DEM enables us to use fine resolution topographic attributes over large spatial areas for understanding micro-topographic controls on the land-surface. These relate to impacts of micro-topography on surface runoff paths, moisture availability to plants and biota, and the exchange fluxes between surface and subsurface water flow. To address such challenges by taking advantage of extreme-resolution data, a new class of models are required that explicitly resolve processes at high resolution over large areas, including micro-topographic effects, while at the same time maintaining fidelity to governing principles. To this end a modeling system that links MLCan, a vertically resolved model of canopy-root-soil biophysical processes, with GPU-based fully conjunctive surface-subsurface flow model (GCSFlow) has been developed. The MLCan model incorporates both C3 and C4 photosynthetic pathways and resolves the vertical radiation, thermal, and environmental regimes within the canopy. In addition to plant water uptake, hydraulic redistribution is included to simulate the passive transport of moisture throughout the root and soil column following water potential gradients. The GCSFlow model simulates both surface runoff and vertical and lateral transport of moisture belowground. Modifications of the conjunctive flow model for utilizing GPU parallel computing capability significantly enhance the performance of the simulations. Several comparisons with analytical solutions and other numerical models have been carried out for a number of soil properties, flow conditions, and domain geometries to benchmark the physical responses of the conjunctive flow model. An application to a real-world catchment, the Goose Creek watershed located in central Illinois, USA, is also presented to demonstrate the scalability of the parallel, coupled model and its robust performance. The coupled model thus provides more capabilities to address large-scale ecohydrologic modeling problems at extremely high-resolution.