Time‐lapse imaging of soil moisture using electromagnetic conductivity imaging: Wetting phase. Issue 3 (22nd April 2021)

• Record Type: Journal Article
• Authors: Zare, Ehsan; Li, Nan; Arshad, Maryam; Nachimuthu, Gunasekhar; Triantafilis, John
• Is Part Of: Soil Science Society of America Journal. Volume 85:Issue 3(2021)
• Page Start: 760
• Page End: 775
• Publication Date: 2021-04-22
• Subjects: Soils -- United States -- Periodicals; Soil science -- Periodicals; Periodicals; 631.4973
• DOI: 10.1002/saj2.20192
• Languages: English
• ISSNs: 0361-5995

Abstract

Abstract: The Vertosols of the lower Namoi valley in New South Wales, Australia, have been extensively developed for irrigated cotton ( Gossypium hirsutum L.) production. However, there are competing demands for water from mining, environmental needs, and agriculture, with climate change forecasts suggesting reduced rainfall and hotter climates. This means irrigators need to improve water efficiency, which requires methods to monitor the soil volumetric water content (θ, m 3 m –3 ). The aim of this study is to assess if a multicoil electromagnetic instrument (DUALEM‐421), which measures soil apparent electrical conductivity (ECa, mS m –1 ), can be used to value add to limited measurements of θ and allow extrapolation along a uniformly heavy clay irrigation furrow. Specifically, we seek to invert ECa using EM4Soil software to generate electromagnetic conductivity images, whereby the estimates of soil true electrical conductivity (σ, mS m –1 ) in the topsoil, subsurface, and subsoil are correlated with θ measured by Decagon GS3 sensors installed at the same depths. We also compared the usefulness of inversion of different combinations of DUALEM ECa : DUALEM‐1, DUALEM‐2, and DUALEM‐21. We found estimated σ from inversion of DUALEM‐421 ECa had good agreement (i.e., Lin's concordance = .84) with measured soil bulk electrical conductivity (σb, mS m –1 ). We explored the relationship between DUALEM‐421 estimated σ and θ using an artificial neural network. The predictions showed the model had good coefficient of determination ( R 2 = .74) with good accuracy (RMSE = 0.04 m 3 m –3 ) and demonstrated good agreement between measured and predicted θ (Lin's = .84). Core Ideas: Proximal ancillary data were used to map soil volumetric water content (θ, m 3 m 3 ). Inversion software was used to estimate true electrical conductivity (σ) from DUALEM‐421 ECa . Time‐lapse images of θ were created during wetting phase.