Mapping cation exchange capacity using a quasi-3d joint-inversion of EM38 and EM31 data

Event type: 
18 July 2019

Rountree Room, L3 Biolink

Dongxue Zhao

The cation exchange capacity (CEC, cmol(+)/kg) is a measure of soil’s capacity to retain exchangeable cations. However, it is expensive to collect CEC across a heterogenous field and at different depths. To value-add to limited data, proximal sensed electromagnetic (EM) data has been coupled to CEC through linear regression (LR) models, because they measure apparent soil electrical conductivity (ECa, mS/m). However, these LRs have been depth-specific. This approach was compared with one universal LR between estimates of true electrical conductivity (s, mS/m) and CEC from various depths, including topsoil (0-0.3 m), subsurface (0.3-0.6 m), shallow subsoil (0.6-0.9 m) and deeper subsoil (0.9-2.1 m). We estimated s from inversion of EM38 and EM31 ECa either alone or in combination (joint-inversion), in horizontal (ECah) and vertical (ECav) modes, using a quasi-3d (q3-d) inversion software (EM4Soil) and various parameters, including EM38 at two different heights (i.e. 0.2 or 0.4 m). In terms of performance, the LR correlation (R2 > 0.60) was largest between deeper subsoil CEC and EM38 ECah at 0.2 m. However, the LR was unsatisfactory for CEC calibration in the topsoil (0.31), subsurface (0.37) and shallow subsoil (0.52). In comparison, a universal LR between CEC and σ was well correlated (0.72), when both EM38 (0.2 m) and EM31 ECa in both modes, were inverted using a forward model (CF), inversion algorithm (S2) and small damping factor (λ = 0.03). The calibrations tested using a leave-one-out cross validation, showed CEC prediction was precise (RMSE, 2.35 cmol(+)/kg), unbiased (ME, -0.002 cmol(+)/kg) with good concordance (Lin’s, 0.83). To improve areal prediction, closer spaced transects need to be collected, while improved vertical resolution of CEC prediction we recommend DUALEM-421 ECa data be acquired. 

Bio: Dongxue Zhao is a second year PhD student of #UNSWSoilScienceCentral2019. Her research focuses on application of proximal and remote sensors to develop digital soil maps. Specifically, her interests include establishing a visible near infrared spectral library to predict soil physical and chemical properties and mapping cation exchange capacity in three dimensions using an electromagnetic inversion algorithm.