QUANTIFYING WOOD MOISTURE CONTENT USING 3D GROUND PENETRATING RADAR AND ELECTRICAL RESISTIVITY TOMOGRAPHY

Abstract

Sap flow measurement using heat sensors is popular technique for measuring tree water use. However, these measurements are local and their accuracy depends largely on the accuracy of estimating the entire sapwood area. Ground Penetrating Radar (GPR) and Electrical resistivity Tomography (ERT) are suitable technologies for mapping moisture content variations, since both methods are highly sensitive to changes in water saturation. Here we describe a coupled ERT and GPR data processing approach to estimate moisture content within tree trunks. The ERT results are used to infer the depth of the interface between the water-filled sapwood and dry heartwood areas, from which an estimate of the radar wave velocity within the sapwood and heartwood areas are determined using the corresponding arrival times in the GPR results. Dielectric permittivity is then calculated from the radar velocity, and finally moisture content is estimated using an empirical relationship (Topp equation). In addition, the GPR direct waves are used to estimate the moisture content along the outer boundary of the tree trunks. We applied our approach on field-based GPR and ERT data collected on lodgepole pine trees around Laramie (WY). The ERT data were collected using mixed array configurations to obtain high vertical and horizontal resolutions, and the GPR data was collected using 800 MHz shielded antenna to ensure high vertical resolution. To eliminate reflections from features outside tree volume, we surrounded both the GPR antenna and the tree trunk with an electromagnetic shielding sheet during data acquisition. To ensure obtaining reliable models, we apply a joint GPR and ERT inversion approach that allows for structural similarities only in the areas within the model space where both images show similar variations. We demonstrate that our approach facilitates monitoring and estimating tree water use.