Relating soil salinity, clay content and water vapour sorption isotherms

• Published in: European journal of soil science Vol. 71; no. 3; pp. 399 - 414
• Main Authors: Chen, Mingyuan, Zeng, Wenzhi, Arthur, Emmanuel, Gaiser, Thomas, Lei, Guoqing, Zha, Yuanyuan, Ao, Chang, Fang, Yuanhao, Wu, Jingwei, Huang, Jiesheng
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01-05-2020; Wiley Subscription Services, Inc
• Subjects: Clay; clay fraction; Clay soils; DLP model; Electrical conductivity; Electrical resistivity; geometric characteristics; Hysteresis; Identification; isotherm models; Isotherms; Levels; Moisture content; Polynomials; Prediction models; Properties; Regression analysis; Robustness; Saline soils; Salinity; Salinity effects; Salt; Salts; Shape; Soil; Soil investigations; Soil salinity; Soil water; Sorption; Vapor pressure; Vapour pressure; Water vapor; Water vapour
• ISSN: 1351-0754; 1365-2389
• DOI: 10.1111/ejss.12876

An accurate description of soil water vapour sorption isotherms (WSIs) is significant for modelling numerous soil processes. Soil clay content predictions from WSIs can be cost effective and convenient, although the approach has only been investigated on non‐saline soils. To investigate relationships among the clay content, water vapour sorption characteristics and salinity of soils, the WSIs of 35 soil samples were measured at four salinity levels. The characterization of the isotherms indicated that the double log polynomial (DLP) model was optimal for fitting the WSIs of soils with various salinity levels. Moreover, existing clay content prediction models based on WSIs had poor applicability to saline soils, and soil salinity had a greater influence than the soil clay content on the shape of the WSIs. Therefore, multiple linear regression (MLR) models were established to explain the relationships among the soil clay content, salt content and geometrical characteristics of the WSIs. Specifically, a significant correlation was observed between the soil salt content and the sub‐area at the end interval of the WSIs (relative vapour pressure (RH) = 0.83 to 0.93, coefficient of determination (R2) > 0.9) and the hysteresis sub‐area in the middle of the WSIs (RH = 0.53 to 0.63, R2 > 0.8). Furthermore, a strong linear relationship existed between the clay content and the sub‐areas of desorption and hysteresis at the end interval of the WSIs (RH = 0.73 to 0.83, R2 > 0.9) in the non‐saline soils only. Based on the investigated samples, the geometrical characteristics of WSIs were affected predominantly by the soil clay content when the salt content was low (electrical conductivity (EC) < 4.5 dS m−1), whereas salinity was the main influencing factor for samples that have high salt content (EC ≥ 4.5 dS m−1). Moreover, the WSIs' geometric characteristics had the potential to predict soil salt content and clay content under certain limited conditions. Isotherm models that accurately describe water sorption isotherms of saline soils were identified. Soil clay content and salinity affect water sorption isotherms differently depending on soil salinity level. Robust relationships between the soil salinity, clay content and WSI shape properties were established.
Highlights
Isotherm models that accurately describe water sorption isotherms of saline soils were identified.
Soil clay content and salinity affect water sorption isotherms differently depending on soil salinity level.
Robust relationships between the soil salinity, clay content and WSI shape properties were established.