Reliability of acoustic tomography and ground-penetrating radar for tree decay detection

Various nondestructive testing technologies have been used for detecting and visualizing internal defects in urban trees. The results obtained by using different nondestructive testing tools can be interpreted in different ways, which may result in inaccurate assessment of the true condition of the inspected trees. The main objective of this study was to evaluate the reliability of acoustic tomography and ground-penetrating radar (GPR) technology for detecting internal decay in a number of different tree species.

One hundred and forty-seven individual trees of 33 species were inspected at a historic park in Yangzhou, Jiangsu Province, China, using a combination of visual inspection, acoustic tomography, GPR scanning, and resistance micro-drilling methods. Special image processing procedures were developed to analyze the acoustic and radar images and to estimate the proportion of compromised wood.

The acoustic tomography tests revealed 10 defective trees with acoustic shadows suggesting compromised wood in more than 10% of the cross-section area. The actual nature of these defects on living trees can be ascertained by conducting resistance micro-drilling at selected paths. The Tree Radar Unit (TRU) System produced 85% false positive predictions and thus was not successful in visualizing the true physical conditions of the trees.

Acoustic tomography can successfully detect trees with internal decay and cavities. A combination of visual inspection, acoustic tomography, and resistance micro-drilling is an effective approach to detect, measure, and visualize internal defects on a diversity of tree species. The image processing procedures we developed make possible the quantitative analysis of compromised wood and could improve the science-based tree risk assessment process. In contrast, the TRU System presented challenges in field applications, particularly on trees with small diameters and irregularly shaped trunks. The images obtained in this study using the TRU system were largely inaccurate and not reliable for tree inspection.