Mapping Infiltration Shape Parameters for Enhanced Soil Hydraulic Characterization

Purpose. Accurate characterization of water infiltration into the vadose zone requires estimating key soil hydrodynamic properties, including the macroscopic capillary length (λ_c) and saturated hydraulic conductivity (K_s). These parameters quantify the contributions of capillarity and gravity (λ_c) and gravity-driven flow (K_s) during infiltration. Both λ_c and K_s can be estimated in the field using a simple Beerkan infiltration test, which requires minimal equipment, limited water, and no specialized operators. Their calculation, however, depends on “integral” shape parameters that vary strongly with soil type. In this study, we present new formulations based on integrating the hydraulic conductivity function expressed in terms of pressure head. These formulations allow accurate estimation of shape parameters even under dry soil conditions, providing an alternative to previous methods that rely on diffusivity or conductivity expressed as a function of water content.

Method. We applied the new formulations to calculate soil-dependent shape parameters for the twelve USDA textural classes. Their performance in estimating λ_c and K_s was evaluated using synthetic cumulative infiltration curves generated with HYDRUS-2D/3D and compared with results obtained using default literature values. For practical applications, we propose two approaches to select appropriate shape parameters: (i) based on soil samples to determine textural class, and (ii) using texture-dependent parameter maps for site-specific selection. Both approaches were tested using a dataset of 167 Beerkan infiltration experiments across seven sites in Burundi, Ghana, Italy, and Senegal.

Results. The sample-based approach provides higher accuracy in estimating λ_c and K_s, whereas the map-based approach eliminates the need for laboratory analysis and still outperforms default literature values, making it suitable for large-scale studies. To support the map-based method, we provide parameter maps at 250 m resolution for six countries (Burundi, Cameroon, Ghana, Italy, Kenya, and Senegal), alongside complementary soil property maps from the SoilGrids database (clay, sand, silt content, dry bulk density, and USDA soil texture classes), all freely accessible. Additionally, we propose a simplified method for estimating λ_c using a new empirical relationship that requires only the Mualem–van Genuchten shape parameter n, which can be derived from SoilGrids texture data using pedotransfer functions such as Rosetta3. Complementary maps of all Mualem–van Genuchten parameters are also provided.

Conclusions. This work improves and simplifies the field estimation of key soil hydrodynamic properties by providing shape parameters for all USDA texture classes and accessible maps for parameter extraction. The approach facilitates the hydraulic characterization of large areas and extensive datasets, supporting both local and regional-scale infiltration studies.

Data availability

Soil property and parameter maps at 250 m resolution for Burundi, Cameroon, Ghana, Italy, Kenya, and Senegal are available in the open-access digital repository Zenodo at https://doi.org/10.5281/zenodo.17397791

Funding

This work was supported through the project GALILEO ― Strengthening rural livelihoods and resilience to climate change in Africa: innovative agroforestry integrating people, trees, crops and livestock (project number: 101181623), funded by the European Union.