Per Loll

Predicting saturated hydraulic conductivity from air permeability: Application in stochastic water infiltration modeling

Several relationships exist for predicting unsaturated hydraulic conductivity K(ψ) from saturated hydraulic conductivity Ks and the soil-water retention curve. These relationships are convenient for modeling of field scale system sensitivity to spatial variability in K(ψ) . It is, however, faster and simpler to measure air permeability ka at ψ = −100 cm H2O, than Ks. This study explores the existence of a general prediction relationship between ka, measured at −100 cm H2O, and Ks. Comparative analyses between ka-Ks relationships for nine Danish and Norwegian soils, six different soil treatments, and three horizons validated the establishment of a soil type, soil treatment, and depth/horizon independent log-log linear ka-Ks relationship. The general ka-Ks relationship is based on data from a total of 1614 undisturbed, 100-cm3 core samples and displays general prediction accuracy better than ±0.7 orders of magnitude. The accuracy and usefulness of the general relationship was evaluated through stochastic analyses of field scale infiltration and ponding during a rainstorm event. These analyses showed possible prediction bias associated with the general ka-Ks relationship, but also revealed that sampling uncertainty associated with estimation of field scale variability in Ks from a limited number of samples could easily be larger than the possible prediction bias.

Air and water permeability in differently textured soils at two measurement scales

Air permeability can be used to describe the structure of the soil but may also be used to predict saturated hydraulic conductivity. This raises the question of whether the two parameters exhibit the same degree of scale dependency. In this study the scale dependency of water permeability (saturated hydraulic conductivity, Kw) and air permeability (ka, at a matric water potential of −50 cm H2O) was tested at four different sites (three horizons at each site), by using two measurement scales (100 cm3 and 6280 cm3). No clear effect of scale on variability was observed. Air and water permeability displayed higher variabilities for two structured loamy soils compared with two sandy soils. For the more structured soils, the variability between measurements was lower for air compared with water permeability. Both air and water permeabilities were higher at the large scale compared with the small scale, but this scale-dependent difference was less pronounced in sandy soils, suggesting a smaller representative elementary volume. For three of the four soils, a highly correlated relationship between Kw and ka on both small and large soil samples was observed. For the fourth soil, water retention data revealed that the samples were not sufficiently drained at −50 cm H2O to validate a comparison between the two parameters. Predictive Kw (ka) relations for the remaining three soils at the two scales compared favorably with a general Kw (ka) relation proposed by Loll et al. (1999). This study supports the use of a general predictive relation between ka near field capacity (at around −50 to −100 cm H2O) and Kw, but caution should be taken if the soil has a large content of pores that will drain at or close to a matric water potential of −50 cm H2O.

A new two‐step stochastic modeling approach: Application to water transport in a spatially variable unsaturated soil

A new two-step stochastic modeling approach based on stochastic parameter inputs to a deterministic model system is presented. Step I combines a Stratified sampling scheme with a deterministic model to establish a deterministic response surface (DRS). Step II combines a Monte Carlo sampling scheme with the DRS to establish the stochastic model response. The new two-step approach is demonstrated on a one-dimensional unsaturated water flow problem at field scale with a dynamic surface flux and two spatially variable and interdependent parameters: The Campbell [1974] soil water retention parameter (b) and the saturated hydraulic conductivity (Ks). The new two-step stochastic modeling approach provides a highly time efficient way to analyze consequences of uncertainties in stochastic parameter input at field scale. The new two-step approach is competitive in analyzing problems with time consuming deterministic model runs where the stochastic problem can be adequately described by up to two spatially variable parameters.

Stochastic analyses of field‐scale pesticide leaching risk as influenced by spatial variability in physical and biochemical parameters

Field-scale pesticide leaching risk assessments were performed by incorporating a numerical, one-dimensional, water and pesticide transport and fate model into the two-step stochastic modeling approach by Loll and Moldrup [1998]. The numerical model included first-order pesticide degradation, linear equilibrium adsorption, and plant uptake of water and pesticide. Simazine was used as a model pesticide, and leaching risk was expressed as the cumulative mass fraction of applied pesticide leached below 100 cm after 1 year. Spatial variability in soil physical and biochemical data, as well as measured meteorological data from an average and a relatively wet year, was considered for two Danish field sites: (1) a coarse sandy soil, with relatively small variability in hydraulic properties, and (2) a sandy loam, with large variability in hydraulic properties. The two-step stochastic modeling approach was used to investigate the relative impact of spatial variability in saturated hydraulic conductivity Ks, soil-water retention through the Campbell [974] soil-water retention parameter b, and pesticide sorption through the organic carbon content (OC). For the coarse sandy soil, field-scale spatial variability in OC was the single most important parameter influencing leaching risk, whereas for the sandy loam, Ks was found more important than OC. The relative impact of field-scale spatial variability in these parameters was found independent of the meteorological conditions, whereas the absolute level of leaching risk was highly dependent on the meteorological conditions. Assuming a linear dependency between pesticide half-life and OC, a unified approach to modeling simultaneous field-scale variability in biodegradation and adsorption was proposed. Leaching risk assessments based on this approach showed that the parts of the field with both low biological activity and low adsorption capacity contributed with a dramatic increase in leaching risk, and suggested that field-scale spatial variability in biochemical processes can be of similar or larger importance than both hydraulic properties and meteorological conditions.