Influence of saturated water content on estimating soil hydraulic properties from cumulative disc infiltrometer measurements
Résumé
The soil sorptivity, S, and saturated hydraulic conductivity, K s , are fundamental soil hydraulic properties that can be estimated from the cumulative infiltration curve measured with a disc infiltrometer. The Haverkamp infiltration model is widely used to estimate S and K s . This model includes as inputs the constants β and γ and the difference between the initial, θ i , and final, θ s , volumetric water contents, Δθ. Since Δθ would be expressive of the possible measurement errors, and assuming β, γ, and θ i as known values, the first objective of this work is to analyze the influence of θ s on the optimization of K s and S. To this end, a sensitivity analysis, which consists of estimating K s and S for a range of θ s was applied on synthetic infiltration curves simulated for homogeneous columns of sand and loam soil. Then, and working on real soils under different tillage management, we evaluated different procedures to measure θ s and analyzed its impact on K s and S estimation. Four different techniques were compared: the gravimetric-core method and two TDR invasive (3 and 5 cm) and a non-invasive (NiP) probes. All TDR probes were connected to a low-cost NanoVNA. The sensitivity analysis showed that θ s , K s and S can be optimized simultaneously from the inverse analysis of an infiltration curve when β and γ are known values and the infiltration curve is near the steady-state zone. However, due to the intrinsic complexities of real soils and the fact that β and γ are unknown variables, we recommended to optimize K s and S using measured θ s . The NiP sensor connected to a NanoVNA provided a fast, inexpensive, clean, accurate and robust alternative to measure θ s at the end of the infiltration experiments.
Origine | Fichiers éditeurs autorisés sur une archive ouverte |
---|---|
licence |