Skip to content

Hodnett & Tomasella (2002), tropical-soil regressions for four van Genuchten parameters.

Source

Hodnett, M. G., & Tomasella, J. (2002). Marked differences between van Genuchten soil water-retention parameters for temperate and tropical soils: A new water-retention pedo-transfer functions developed for tropical soils. Geoderma, 108(3-4), 155-180.

DOI: 10.1016/S0016-7061(02)00105-2

Scope

Territory: Tropical soils between approximately 25 degrees N and 25 degrees S.

Dataset: The IGBP-DIS tropical-soil dataset contains 771 retained horizons from 249 profiles in 22 countries, split into 492 calibration curves and 279 validation curves.

Functions

calc_ptf_hodnett2002

Estimate four van Genuchten water-retention parameters for tropical soils.

Status: implemented

Prediction target: Residual and saturated volumetric water content and the alpha and n shape parameters of the van Genuchten water-retention model.

Models: \(h(\theta)\) — van Genuchten model with m = 1 - 1 / n

Inputs

Name Type Unit Domain Description
sand number % 0.5 <= value <= 99.0 Sand content in the 0.05-2 mm USDA particle-size fraction, expressed as a percentage.
silt number % 0.4 <= value <= 77.4 Silt content in the 0.002-0.05 mm USDA particle-size fraction, expressed as a percentage.
clay number % 0.0 <= value <= 95.4 Clay content in the less-than-0.002 mm USDA particle-size fraction, expressed as a percentage.
organic_carbon number % 0.0 <= value <= 30.8 Soil organic carbon content expressed as a percentage.
bulk_density number Mg/m^3 0.28 <= value <= 1.88 Soil bulk density.
cation_exchange_capacity number cmol/kg 0.0 <= value <= 93.7 Soil cation exchange capacity.
ph number dimensionless 3.60 <= value <= 9.60 Soil pH.

Outputs

Name Unit Domain Description
alpha kPa^-1 value > 0 Shape parameter of the van Genuchten water-retention model.
n dimensionless value > 0 Shape parameter controlling water-retention curve steepness.
theta_s m^3/m^3 Saturated volumetric soil water content.
theta_r m^3/m^3 Residual volumetric soil water content.

Note

The regressions use percentages of sand, silt, clay, and organic carbon, bulk density in Mg/m^3, CEC in cmol/kg, and pH.

Note

Table 8 reports the reoptimised Step 3 coefficients used here and states that its displayed values are multiplied by 100; each polynomial is therefore divided by 100 before alpha and n are back-transformed.

Note

The fitted ln(alpha) and ln(n) values are back-transformed with the natural exponential, as specified by the paper's use of natural logarithms.

Warning

The paper reports n = 497 when introducing the continuous PTF, but its dataset split gives 492 calibration curves and 279 validation curves, which account for all 771 retained horizons. The extraction treats 497 as a typographical error and records 492 as the calibration-set size.

Warning

The PTF did not reproduce water-retention curves accurately for soils with bulk density below 0.8 Mg/m^3, most of which were Andosols.

Warning

The PTF did not reproduce very low alpha values well because high-alpha soils dominated the calibration data; mineralogy and structure were not directly represented.