Investigating scale effects in soil water retention curve via spatial time domain reflectometry

The soil water retention curve (SWRC) is usually measured by the axis translation technique (ATT), assuming a specimen thickness of a few centimetres. Usually, this assumption is unquestionable to ensure the representative elementary volume (REV). However, such an assumption cannot always be conserved with coarser soil regarding the local heterogeneities generated by larger pore sizes. In addition, the authentic SWRC is supposed to be the soil moisture profile above the phreatic surface. Those issues, therefore, raise concerns about the scale effects in SWRC. With an aim to investigate them, this work compared the primary drainage SWRCs of coarse, medium and fine sand measured by the standard hanging column and full-scale soil column (150 cm). The soil moisture profile at the final drainage stage was detected using the spatial time domain reflectometry (spatial TDR) technique with a presentation of sensor calibrations and validations. The observations showed discrepancies between SWRCs given by two methods in terms of air entry value (AEV) and slope of SWRC but agreements on residual moisture content. Moreover, such differences could be alleviated with the soil becoming finer and better graded. Finally, the physical reasons are discussed, referring to local heterogeneity and pressure gradient changes on the top boundary. In conclusion, the full-scale soil column test can be utilised to avoid scale effects, despite demanding high costs and massive labour efforts. Also, a shorter soil column incorporating point-wise suction and moisture sensors in prior studies could be an alternative solution.