JOINT INVERSION OF MULTI-CONFIGURATION ELECTROMAGNETIC INDUCTION MEASUREMENTS TO PREDICT SOIL WETTING PATTERNS DURING SURFACE TRICKLE IRRIGATION

Abstract

Spatial and temporal distribution of soil moisture pattern under trickle emitter is a pre-requisite for the design and management of efficient trickle-irrigation system. Wetting front dimensions are usually obtained by direct measurement of soil wetting in field, which is time consuming and provide local information. Furthermore, mathematical models are used to describe infiltration from a point source to design and manage trickle systems. However, several simplified assumptions in the models affect the utility of numerical models. In this respect, non-invasive geophysical methods such as low frequency electromagnetic induction (EMI) system is becoming a powerful tool to map spatial and temporal soil moisture patterns due to fast measurement capability and sensitivity to soil water content and salinity. In this study, a new electromagnetic system, the CMD mini-Explorer, is used for soil characterization to measure the wetting patterns of trickle-irrigation system using joint inversion of multi-configuration EMI measurements. Six transects of EMI measurements are carried out in a farm where Acacia trees are irrigated with brackish water using a drip irrigation system. Global and local optimization algorithms are used sequentially, to minimize the misfit between the measured and modeled apparent electrical conductivity to reconstruct vertical electrical conductivity profile. The electromagnetic forward model based on full solution of the Maxwell's equation is used as the EMI measurements are performed under high induction number conditions. Inversion of calibrated measured data for the first transect using different instrument orientations and offsets provides lateral and vertical conductivity variations very similar to those observed in ground-truth measurement. High values of soil electrical conductivity in the inversion results show the dimension of wetting front of brackish water. The proposed approach allows for quantitative mapping and monitoring of the spatial electrical conductivity variations and can be utilized for proximal sensing of the subsurface properties with wide variety of applications.