Exchange of water between the atmosphere and biosphere via evapotranspiration influences global hydrological, energy, and biogeochemical cycles. Isotopic analysis has shown that evapotranspiration over the continents is largely dominated by transpiration. Water is taken up from soil by plant roots, transported through the plant’s vascular system, and evaporated from the leaves. Yet many Land Surface Models (LSMs) integrated into Earth System Models (ESMs) treat plant roots as passive components. These models distribute the ET sink vertically over the soil column, neglect the vertical pressure distribution along the plant vascular system, and assume that leaves can directly access water from any soil layer within the root zone. Numerous studies have suggested that increased warming due to climate change will lead to drought and heat-induced tree mortality. A more mechanistic treatment of water dynamics in the soil-plant continuum (SPC) is essential for investigating the fate of ecosystems under a warmer climate. A range of site-level SPC model structures have been developed recently (e.g., resolving transport of water through rhizosphere, including radial and axial transport through roots, and accounting for plant water storage) while the appropriate model structure for global LSMs remains unknown.

In this work, we describe a flexible SPC modeling framework that we have coupled to the E3SM Land Model (ELM). The SPC model uses the variably saturated Richards equations to simulate water transport. The model uses individual governing equations and constitutive relationships for the various SPC components (i.e., soil, root, and xylem). Finite volume and backward Euler discretizations are used to solve the SPC model and the Portable, Extensible Toolkit for Scientific Computation (PETSc) provides the solution of discretized system of equations. PETSc’s multi-physics coupling capability via DMComposite provides SPC modeling framework with a flexibility to represent a range of model fidelity. Verification of the SPC modeling framework is performed using Method of Manufactured Solutions. Numerical results are presented for the application of the SPC model to study sites in Silas Little Experimental Forest, NJ and the Ameriflux-affiliated eddy covariance tower at University of Michigan Biological Station.