Land Modeling Overview#
Land models are used with atmospheric circulation models or coupled Earth System Models in order to provide the lower boundary conditions to the atmosphere. As such, the exchange of heat, moisture and momentum at the interface between the land surface and the atmospheric boundary layer is strongly linked to the vertical turbulent diffusion parameterization in the atmospheric physics. The surface fluxes are computed by solving the energy balance at the surface and by describing the heat and water dynamics in the sub-surface soil. In setups including an ocean, river discharge resulting from the surface runoff and sub-surface drainage is also coupled to the ocean as freshwater flux from the land.
There are two atmospheric physics packages in ICON, ICON-NWP and ICON-AES. Historically, ICON-NWP contained the “turbdiff/turbtran” turbulence parameterization with the TERRA land model and ICON-AES the “vdiff” turbulence parameterization with the ICON-Land component implementing the JSBACH land model. Recently, the QUINCY land model has been implemented in ICON-Land as an additional configuration option. The following configurations are now possible:
to be added: TMX turbulence package, biogeochemical cycles
Land Parameterizations#
TERRA#
The multi-layer soil-vegetation-atmosphere-transfer component TERRA can be classified as a second-generation land-surface model with the following characteristics:
Infiltration and a multi-layer model for hydrology (Richards equation). The transport of soil water is solved implicitely.
Surface and ground-water runoff
Multi-layer heat conduction equation (NWP: 7-layer + 1 climate layer; 1 cm - 14.58 m). The solution of the heat transport is solved implicitly.
Fractional freezing/melting
Evapotranspiration with bare soil evaporation after Dickinson 1984 and Schulz & Vogel 2020, and transpiration following Jarvis 1976
One-layer snow scheme
Skin-layer approach to mimic canopy effects (Viterbo & Beljaars 1995, Schulz & Vogel 2020)
From a numerical point of view TERRA is coupled explicity to the atmosphere. TERRA requires a set of external physiographic parameters that hold the information such as e.g. soil type or land-use type. To address the issue of spatial soil heterogeneity ICON employs tiles. Tiling is done in land-use space, and TERRA is called separately for each tile. Resulting fluxes from the separate tiles are averaged at a blending height of 10 m and passed to the turbulence scheme.
See also Section 3.8.11 of the ICON Tutorial 2024.
ICON-Land#
ICON-Land is a framework for the modeling of land processes in ICON and can be used as a stand-alone land surface model as well as in the fully coupled ICON Earth System model or in the ICON atmosphere-only model. It is specifically designed in a modular way for the integration of concurrent and alternative process and surface descriptions in a flexible and easy-to-use way. Currently, specific process implementations include the JSBACH and the QUINCY model configurations.
The ICON-Land framework has been designed to systematically separate the infrastructure necessary to implement physical, biogeophysical and biogeochemical land processes from the concrete process implementations which are accessed by abstract interfaces. A further goal was the ability to support different experimental configurations of varying scope and complexity to be used in different global, regional or single-site applications, coupled online to an atmosphere model or driven offline by atmospheric observations.
ICON-Land is implemented in an object-oriented and modular way in Fortran2003/2008. Hierarchical trees are used for the flexible description of surface characteristics (tiles). Scientific code (processes) is clearly separated from the infrastructure. ICON-Land is a self-contained package while using only a basic infrastructure for I/O, parallel domain decomposition, time control, etc. from the ICON model via a relatively small number of adapter routines.
JSBACH#
JSBACH in ICON-Land (aka JSBACHv4) was initially a re-implementation of JSBACHv3 in the new framework where JSBACHv3 Reick et al. 2021 has been the land component of the MPI-M ECHAM and MPI-ESM models used in many modeling studies over the last decades. JSBACHv4 in ICON-Land includes the fast physical and biogeophysical processes as well as biogeochemical processes to simulate the natural land carbon cycle, disturbances and anthropogenic and natural land cover change.
The surface energy balance and the soil thermal layers on land are coupled implicitly to the vdiff vertical diffusion scheme. Plant productivity by photosynthesis, phenology (leaf area index), roughness lengths for momentum and heat, and visible and near-infrared albedos are computed on a flexible number of tiles representing sub-grid surface heterogeneity by plant functional types. Heat and water dynamics in the soil are described to a depth of about 10 m and are coupled to calculations of surface runoff and sub-surface drainage and the resulting river discharge through a hydrologic discharge model (Hagemann & Duemenil 1997, Riddick et al. 2018). A five-layer snow model is applied and the soil dynamics include freezing water and phase changes between liquid and frozen water (Ekici et al. 2014, de Vrese et al. 2021).
QUINCY#
QUINCY (QUantifying Interactions between terrestrial Nutrient CYcles and the climate system, Thum et al. 2019) has been developed at the MPI for Biogeochemistry as an alternative, more comprehensive representation of biogeochemical processes for ICON-Land. QUINCY is integrated with ICON-Land through a set of defined interfaces that affect the biological control of the land surface’s radiation balance, momentum and heat transfer as well as its water balance. It follows a modular design that allows to integrate fundamental, fast biospheric processes with increasing degree of complexity from a prescribed phenology mode, a vegetation dynamics mode to a fully coupled BGC vegetation-soil model. QUINCY also serves as test-bed for novel model components, e.g. representations of soil biogeochemical processes (Yu et al. 2020a, Yu et al. 2020b).
QUINCY simulates the element cycling of carbon, nitrogen, and phosphorus in terrestrial ecosystems, and includes novel representations of source/sink limited plant growth; the acclimation of many ecophysiological processes; an explicit representation of vertical soil processes to separate litter and soil organic matter dynamics; a range of new diagnostics (leaf chlorophyll; isotope tracers) to allow for a more in-depth model evaluation. Case-studies include the evaluation against standard biosphere benchmarks Thum et al. 2019, in particular ecosystem manipulation experiments Caldararu et al. 2020 as well as long-term monitoring data Caldararu et al. 2021.
Dynamic Vegetation#
to be added
Carbon Cycle#
to be added