Monitoring scale-specific and temporal variation in electromagnetic conductivity images

Abstract

In semiarid and arid landscapes, irrigation sustains agricultural activity but because of increasing demands on water resources there is a need to make gains in efficiency. As such spatial variation of soil properties such as clay and salinity needs to be understood because they strongly influence soil moisture availability. One way is to use electromagnetic induction because apparent soil electrical conductivity (ECa) is related to volumetric soil moisture (θ), clay and salinity (ECe). However, depth-specific variation has not been explored. Our aim is to generate electromagnetic conductivity images (EMCIs) by inverting DUALEM-421 ECa and show how true electrical conductivity (σ) can be correlated with θ, clay, ECe and bulk density (ρ) on different days post-irrigation (i.e., 1, 4 and 12 days). Two-dimensional multi-resolution analysis (MRA) is used to show how spatio-temporal variation in σ is scale-specific and how soil properties influence σ at different scales. We study this beneath a pivot irrigated alfalfa crop. We found that σ on days 1 and 4 was correlated with θ (Pearson’s r = 0.79 and 0.61) and clay (0.86 and 0.80) and the dominant scale of variation occurred at 9.3–18.7 m (50.21 % of total variation), >74.7 m (23.18 %) and 4.7–9.3 m (16.29 %). Between 9.3–18.7 and 4.7–9.3 m the variation may be a function of the cutter width (8 m), while >74.7 m may be change in clay and ECe and gantry spacing (~48 m). The sprinkler spacing (1.2 and 1.6 m) explains short-scale variation at 1.2–2.3 m.