Estimating the macroscopic capillary length from Beerkan infiltration experiments and its impact on saturated soil hydraulic conductivity predictions

JOURNAL OF HYDROLOGY

卷:589

Article

Di Prima, Simone^1,2  Stewart, Ryan D.³  Castellini, Mirko⁴  Bagarello, Vincenzo⁵  Abou Najm, Majdi R.⁶  Pirastru, Mario¹  Giadrossich, Filippo¹  Iovino, Massimo⁵  Angulo-Jaramillo, Rafael²  Lassabatere, Laurent² 

[1] Univ Sassari, Dept Agr Sci, Viale Italia 39, I-07100 Sassari, Italy

[2] Univ Lyon, Univ Claude Bernard Lyon 1, UMR 5023 LEHNA, ENTPE,CNRS, F-69518 Vaulx En Velin, France

[3] Virginia Polytech Inst & State Univ, Sch Plant & Environm Sci, Blacksburg, VA 24061 USA

[4] Council Agr Res & Econ, Agr & Environm Res Ctr CREA AA, Via Celso Ulpiani 5, I-70125 Bari, Italy

[5] Univ Palermo, Dept Agr Food & Forest Sci, Palermo, Italy

[6] Univ Calif Davis, Dept Land Air & Water Resources, Davis, CA 95616 USA

关键词: Infiltration;    Macroscopic capillary length;    Beerkan;    Ring infiltrometer;    Hydraulic conductivity;   

DOI  :  10.1016/j.jhydrol.2020.125159

来源: Elsevier

【 摘 要 】

The macroscopic capillary length, lambda(c), is a fundamental soil parameter expressing the relative importance of the capillary over gravity forces during water movement in unsaturated soil. In this investigation, we propose a simple field method for estimating lambda(c) using only a single-ring infiltration experiment of the Beerkan type and measurements of initial and saturated soil water contents. We assumed that the intercept of the linear regression fitted to the steady-state portion of the experimental infiltration curve could be used as a reliable predictor of lambda(c). This hypothesis was validated by assessing the proposed calculation approach using both analytical and field data. The analytical validation demonstrated that the proposed method was able to provide reliable lambda(c) estimates over a wide range of soil textural characteristics and initial soil water contents. The field testing was performed on a large database including 433 Beerkan infiltration experiments, with the 99% of the experiments yielding realistic lambda(c) values. The generated lambda(c) values were then used in conjunction with four different methods for estimating saturated soil hydraulic conductivity, K-s. Estimated K-s values were close to those generated by a reference method, with relative error < 25% in nearly all cases. By comparison, assuming constant or soil-dependent lambda(c) values caused relative errors in K-s of up to 600%. Altogether, the proposed method constitutes an easy solution for estimating lambda(c), which can improve our ability to estimate K-s in the field.

【 授权许可】

Free