Introduction

Description

Soil plays a central role in life support on earth, and it controls many of the land-atmosphere interactions. About 40% of the terrestrial precipitation is returned back through the soil-plant-atmosphere continuum (Katul et al. 2012). Nearly half of the annual global biomass production and associated carbon cycle rely on soil processes. The key roles of soil in global water and carbon cycles feature prominently in mediating terrestrial energy balance and climate (Koster et al. 2009). While traditional representation of soil processes at the column or field scales only provide answers for local problems, the representation of such processes in global climate models remains largely sketchy and uncertain (by “soil processes” we consider aspects of infiltration, drainage, water retention, evaporation, plant water uptake, groundwater interactions and rudimentary aspects of nutrients and parameters derived from soil maps). This state of affairs in which a central compartment of Earth system “falls through the cracks” requires a radical change of modeling approaches for providing adequate representation of soil process at regional to global scales where information is needed most (Vereecken et al. 2016). Evidence of effects of soil and land management on local climate have been studied by Pielke (2005), and more recently by Betts et al. (2013) and Davin et al. (2014) that have demonstrated that soil and land-use shape critical aspects of local climate. The nature of such regional feedbacks transcends traditional field scale models of soil processes, yet, these are not yet captured by lumped and undifferentiated soil representations in many land surface models.

Members of the soil/groundwater science community seek a path forward to improve the representation of these processes in land models and to integrate these models as key components of Earth System Models (ESMs) that incorporate all essential processes that govern Earth’s climate. A principal goal of this development is to place soil and subsurface processes in a meaningful spatio-temporal context and with highly resolved soil parameterizations in the core of an advanced global climate and oceanic model, rather than serving as a minimalistic lower (surface) boundary condition as in most models. The aim is to present a new Earth System Model (ESM) built from the soil-vegetation perspective up, rather than from the atmospheric perspective down. The envisioned project would lead to the first comprehensive soil system model (see ISMC: https://soil-modeling.org/governance) capable of local to regional and global process representation, including two-way feedbacks between soil and atmosphere with direct links to groundwater resources. The new capabilities of the modeling and parameterization framework would enable future investigations of globally important science challenges where soil processes play a central role. Investigations with the new ESM would provide new insights into soil-driven processes that influence ecosystem functioning and regional climates, a better understanding of coupled processes that have been obscured by the complexity of terrestrial interactions with climate variations, and the first comprehensive soil systems model scalable from the field to catchment and the globe. The ESM would provide unique educational opportunities via student exchange and interactions across institutions; summer schools and hands-on experiences with methods used in soil, remote sensing, and atmospheric sciences; interactions and learning of new data products (soil, climate, vegetation) and active observatories and platforms.

The Swiss Federal Institute of Technology at Zurich (ETH), Forschungszentrum-Jülich (FZJ), and the International Soil Modeling Consortium (ISMC: https://soil-modeling.org/governance) were interested to fund preliminary developments that would produce a prototype version of the ESM in a minimal amount of time. Through a series of meetings with modelers and other scientists from soil/groundwater, ecosystem, and atmospheric communities, they identified the Ocean-Land-Atmosphere Model (OLAM), developed by Robert Walko and Roni Avissar who are currently at the University of Miami, as an ideal starting point for this development. OLAM includes a global non-hydrostatic atmospheric model for weather and climate simulations that has been developed and applied since the early 2000s (Walko and Avissar 2008a, b, 2011). It is an outgrowth of the Regional Atmospheric Modeling System (RAMS; Cotton et al., 2003), a widely-used research model that has been applied in numerous investigations of mesoscale and cloud-scale phenomena for nearly three decades. OLAM was originally constructed as a global version of RAMS that combined the advantages of regional and global models by utilizing a variable-resolution global gridding system with two-way interaction across scales. Recent versions of OLAM accomplish this with a seamless unstructured grid consisting of hexagonal cells, along with a few pentagons and heptagons. This unification of multiple modeling scales within a single framework enables high-resolution modeling in selected geographic regions of interest while avoiding the problems and inherent limitations of applying lateral boundary conditions in limited-area models. The standard land surface component of OLAM is the Land-Ecosystem-Atmosphere Feedback (LEAF) model (Walko et al. 2000), a Soil-Vegetation-Atmosphere Transfer (SVAT) model that represents energy and mass fluxes at the lower boundary of the atmosphere. In its most basic form, LEAF is a column model consisting of multiple soil/bedrock layers, vegetation, a vegetation canopy air layer, and “surface water” layers to accommodate snow cover or standing water that is either a wetland or has not had time to run off or infiltrate during or following a precipitation event. Water and energy are stored in each component and exchanged between components and the atmosphere, subject to conservation laws, heat conduction and Richards’ equation in the soil, turbulent transfer relationships in the canopy and atmosphere, a stomatal conductance formulation, and radiative transfer. Latent heat energy of freezing and melting are accounted for in all components of LEAF. More complex forms of LEAF are optional in OLAM, including a representation of land-surface biophysics, vegetation dynamics, and soil carbon and nitrogen biogeochemistry via the Ecosystem Demography Version 2 Model (ED2) developed by Medvigy et al (2009).

Walko and Avissar discussed the current formulation of OLAM/LEAF with all parties involved, and specific tasks were identified that would transform the model into the prototype ESM that would form the basis of future research and development. These tasks and an approximate timetable for completion are listed below and constitute a guideline of milestones and deliverables to be accomplished under this project. The project was awarded to the University of Miami, with all tasks to be carried out by Robert Walko, who is the lead developer of OLAM and LEAF. The purpose of this report is to describe the work and tasks that have been completed to date, those that are still in progress, and decisions that have been made during the course of the project.