Spatial Reconstruction of Soil Moisture Content using Non-Contact Thermoacoustic Imaging

Sensing of soil water content is useful in many precision agriculture and resource management applications, particularly if the sensing technique permits frequent field-scale measurements at depth. To date, soil moisture sensing technologies have a trade-off between point based measurements at depth or rapid, remote measurements of water content near the surface. In this paper, we propose a non-contact thermoacoustic soil moisture sensing modality which could permit high-resolution, high-throughput mapping of water content at depth. Within, we develop an algorithm for reconstructing the speed-of-sound in soil, which is known to be highly correlated with the soil moisture content. Through verification in simulation, our algorithm demonstrates high fidelity - reconstructing speed-of-sound profiles that match well with the ground-truth.