Abstract
Abstract. Agroecosystem models, regional and global climate models, and numerical weather prediction models require adequate parameterization of soil hydraulic properties. These properties are fundamental for describing and predicting water and energy exchange processes at the transition zone between solid earth and atmosphere, and regulate evapotranspiration, infiltration and runoff generation. Hydraulic parameters describing the soil water retention (WRC) and hydraulic conductivity (HCC) curves are typically derived from soil texture via pedotransfer functions (PTFs). Resampling of those parameters for specific model grids is typically performed by different aggregation approaches such a spatial averaging and the use of dominant textural properties or soil classes. These aggregation approaches introduce uncertainty, bias and parameter inconsistencies throughout spatial scales due to nonlinear relationships between hydraulic parameters and soil texture. Therefore, we present a method to scale hydraulic parameters to individual model grids and provide a global data set that overcomes the mentioned problems. The approach is based on Miller–Miller scaling in the relaxed form by Warrick, that fits the parameters of the WRC through all sub-grid WRCs to provide an effective parameterization for the grid cell at model resolution; at the same time it preserves the information of sub-grid variability of the water retention curve by deriving local scaling parameters. Based on the Mualem–van Genuchten approach we also derive the unsaturated hydraulic conductivity from the water retention functions, thereby assuming that the local parameters are also valid for this function. In addition, via the Warrick scaling parameter λ, information on global sub-grid scaling variance is given that enables modellers to improve dynamical downscaling of (regional) climate models or to perturb hydraulic parameters for model ensemble output generation. The present analysis is based on the ROSETTA PTF of Schaap et al. (2001) applied to the SoilGrids1km data set of Hengl et al. (2014). The example data set is provided at a global resolution of 0.25° at https://doi.org/10.1594/PANGAEA.870605.
Highlights
Hydraulic properties have fundamental importance in the description of water, energy and carbon exchange processes between the land surface and the atmosphere (e.g. Ek and Cuenca, 1994; Xue et al, 1996)
Several pedotransfer functions (PTFs) have been developed; here, we focus on the widely used PTF ROSETTA model H3 by Schaap et al (2001), which is based on neural network predictions for the estimation of the Mualem–van Genuchten (MvG) parameters θs, θr, α, n, Ks and L, whereby the water retention curve (WRC) to describe the effective volumetric saturation Se is calculated according to
Reliable soil hydraulic parameterization is important for global climate model predictions, including climate reanalyses and weather forecast models
Summary
Hydraulic properties have fundamental importance in the description of water, energy and carbon exchange processes between the land surface and the atmosphere (e.g. Ek and Cuenca, 1994; Xue et al, 1996). Agroecosystem models (e.g. SWAP; van Dam et al, 2008) and land surface models (LSMs; see below) require adequate parameterization of soil hydraulic properties – i.e. the water retention curve (WRC) and the hydraulic conductivity curve (HCC). These properties regulate the relative magnitude of water balance fluxes such as evapotranspiration, infiltration and surface and sub-surface runoff (Vereecken et al, 2016), and the amount of water held in the soil at any one time. With regard to the carbon balance, photosynthesis and soil respiration both strongly depend on soil moisture content and implicitly on choice of soil hydraulic models and their parameters
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have