Quantitative mapping of soil salinity using the DUALEM‐21S instrument and EM inversion software. (16th May 2018)

• Record Type: Journal Article
• Authors: Koganti, Triven; Narjary, Bhaskar; Zare, Ehsan; Pathan, Aslam Latif; Huang, Jingyi; Triantafilis, John
• Is Part Of: Land degradation & development. Volume 29:Number 6(2018)
• Page Start: 1768
• Page End: 1781
• Publication Date: 2018-05-16
• Subjects: digital soil mapping (DSM) -- electrical conductivity -- electromagnetic induction -- quasi‐3d inversion -- soil salinity; Land degradation -- Periodicals; Soil conservation -- Periodicals; Reclamation of land -- Periodicals; Land use -- Periodicals; Economic development -- Environmental aspects -- Periodicals; 333.7315
• DOI: 10.1002/ldr.2973
• Languages: English
• ISSNs: 1085-3278

Abstract

Abstract: To generate baseline data for the purpose of monitoring the efficacy of remediation of a degraded landscape, we demonstrate a method for 3‐dimensional mapping of electrical conductivity of saturated soil paste extract (ECe ) across a study field in central Haryana, India. This is achieved by establishing a linear relationship between calculated true electrical conductivity (σ) and laboratory measured ECe at various depths (0–0.3, 0.3–0.6, 0.6–0.9, and 0.9–1.2 m). We estimate σ by inverting DUALEM‐21S apparent electrical conductivity (ECa ) data using a quasi‐3‐dimensional inversion algorithm (EM4Soil‐V302). The best linear relationship (ECe = −11.814 + 0.043 × σ) was achieved using full solution (FS), S1 inversion algorithm, and a damping factor (λ) of 0.6 that had a large coefficient of determination ( R 2 = 0.84). A cross‐validation technique was used to validate the model, and given the high accuracy (RMSE = 8.31 dS m −1 ), small bias (mean error = −0.0628 dS m −1 ), large R 2 = 0.82, and Lin's concordance (0.93), between measured and predicted ECe, we were well able to predict the ECe distribution at all the four depths. However, the predictions made in the topsoil (0–0.3 m) at a few locations were poor due to limited data availability in areas where ECa changed rapidly. In this regard, improvements in prediction can be achieved by collection of ECa in more closely spaced transects, particularly in areas where ECa varies over short spatial scales. Also, equivalent results can be achieved using smaller combinations of ECa data (i.e., DAULEM‐1S, DUALEM‐2S), although with some loss in precision, bias, and concordance.