A Non-Destructive Method to Estimate Leaf Water Content Using the Ka-Band

Abstract

In this paper we demonstrate the use of Ka-band microwave transmission for remote measurement of relative water content (RWC) in leaves. The system, consisting of commercially available horn antennas connected to a vector network analyzer, offers a simple, inexpensive, and low-energy design compared to other non-destructive methods like terahertz spectroscopy. In addition, Ka-band frequency electronic components require less power and can be used in a wider range of environments. Using leaf samples from the Spathiphyllum Wallisii plant, we derived a model that accepts the power transmitted through the leaf as input and outputs an estimate for the RWC and a recommendation on whether or not to water the plant. We verify our model by comparing it to RWC values obtained from a standard method based on weighing the leaf and soaking/dehydrating it. Results indicate that the approach is valid with acceptable model performance. Further work is proposed to improve robustness and accuracy.