Figure 2: Top, from left to right: Soil sampling locations within the Pinios Hydrologic Observatory, soil data from the European soil database and the LUCAS topsoil map. Bottom: Soil texture input datasets of sand, clay, and organic carbon derived from the three data sources.

2.3 The Land Surface Model

2.3.1 The Community Land Model

The Community Land Model v.5 (CLM5) used in this study is the latest version of the land component in the Community Earth System Model (CESM) as described in detail by D M Lawrence et al. [2019]. CLM5 simulates land surface energy fluxes as well as hydrological, biogeophysical, and biogeochemical processes that are driven by atmospheric input variables in combination with soil and vegetation states and characteristics [D Lawrence et al. , 2018]. These processes are simulated on different subgrid units within a grid cell. Subgrid units include (1) the land unit defining the land use category (e.g., vegetated, urban, crop), (2) the column that is represented by 20 soil and 5 bedrock layers and resolves state variables and fluxes of water and energy in the soil, and (3) the patch level capturing biogeophysical and biogeochemical differences between plant functional types (PFTs) (e.g., broadleaf deciduous forest, evergreen shrub, maize, soy). The one-dimensional multilayer vertical water flow in the soil is simulated using a modified Richards equation [Dingman , 2015]. Soil hydraulic parameters for these calculations are derived from pedotransfer functions of sand and clay [Clapp and Hornberger , 1978; Cosby et al. , 1984] and organic properties of the soil [D Lawrence and Slater , 2008]. With version 5 of CLM, a plant hydraulic stress routine was introduced that uses a simple hydraulic framework to model water transport along a water potential gradient from soil via plant to atmosphere [Kennedy et al. , 2019]. The new configuration replaces soil potential with leaf potential as the basis for plant water stress while root water potential is used to drive root water uptake. A new biogeochemistry and crop module, BGC-Crop, enhanced the representation of major crop functional types and land management practices such as irrigation and fertilization. Unlike natural vegetation that competes for water and nutrients, crops operate on separate soil columns that may be irrigated or non-irrigated, thus allowing for differences in land management [D M Lawrence et al. , 2019].
The recent development of CLM5-FruitTree enables the simulation of deciduous fruit trees and associated management practices in CLM5. The main features of the new sub-model include (1) a perennial phenology routine that allows the woody plant parts to remain on the orchard for several years, (2) carbon storage dynamics that enable the regrowth of annual plant parts, (3) an adapted carbon and nitrogen allocation, and (4) the description of typical management practices such as transplanting, pruning, and orchard rotation. Additionally, a new apple plant functional type was parameterized while fertilization and irrigation use the default CLM5 schemes. The complete model development of CLM5-FruitTree is described in Olga Dombrowski et al.[2022].

2.2.3 Irrigation module in CLM5

Irrigation is performed individually over each irrigated soil column and responds dynamically to SM based on a daily check at 6 am. If crop leaf area is non-zero and if the available soil water over a specified irrigation depth \(z_{\text{irrig}}\) (=0.6 m by default) is below a defined threshold, irrigation is triggered. The irrigation amount is based on the SM deficit (𝐷𝑖𝑟𝑟𝑖𝑔) that is calculated over\(z_{\text{irrig}}\):