Sonic and acoustic tomography: how the images are made and what journals report they map
Sonic tomography times sound waves across a trunk and turns the delays into a colour image of the cross-section. Journal studies report close agreement with cut sections, and note that incipient decay might not slow the sound.
Checked against the sources at the bottom of this page on October 15, 2026. Rules, fees and pay change: the source has the last word.
Sonic tomography is described here as the named journals and the Forest Service report publish it. It is not training and does not teach the work. It is not a diagnosis of any tree and does not say what should be done with one.
How the image is made
Gilbert and Smiley's 2004 paper in Arboriculture & Urban Forestry starts from one principle: sound passing through decay moves more slowly than sound through solid wood. Sensors sit on pins driven into the trunk, and each is tapped in turn with a small hammer. Software takes the transit times and the distances between sensors, works out apparent sound velocity, and draws a two-dimensional tomogram of the cross-section. The paper counts 8 to 12 sensors as usual.
A 2020 paper in the Journal of Forest Science by Baláš, Gallo and Kuneš describes the same set-up. The sensors are magnetic and sit on nails driven through the bark, an electronic caliper records the stem shape, and a colour scale shows velocity in relative ranges. Sound velocity correlates with wood density, modulus of elasticity and moisture. Where decay is incipient, a hard decay, velocity might not change at all. Caves, seams and swollen stems can make a tomogram harder to interpret.
Validation against cut sections
Gilbert and Smiley scanned 27 cross-sections from 13 trees at the Bartlett Tree Research Laboratory in North Carolina, then felled the trees and cut sections at the sensor plane. Ten sections had no decay, and the tomograms showed none. In the other 17, the tomogram showed less decay than the section did. The correlation between tomogram and section was r squared 0.94 across all sections and 0.90 where decay was present. In 14 of the 17 the tomogram under-read by an average of 6 percentage points of area, and in three it over-read by 1.
The paper prints two sets of error figures. Its abstract puts false positives at an average of 3% of the area and false negatives at 8%. The results section gives 2% and 9%, with an average accuracy of 89% where decay was present. Two further limits appear in the discussion. Tomogram quality near the bark was lower, and a defect of 5 to 10 cm is quoted as about the smallest the instrument detects.
A 2016 paper in Applications in Plant Sciences by Gilbert and seven co-authors scanned living tropical rainforest trees in Panama with an Argus PiCUS 3. They paired the scans with ImageJ analysis to quantify the share of intact and compromised wood. They report replicable estimates of internal decay and cavities for trunks of varying shape, wood density and bark thickness.
Electrical resistance as a second view
The Czech paper adds a second instrument that shares the same nails. An electrical resistance tomograph maps resistivity, which follows wood moisture and ion content, and the authors report it can pick up decay before sound velocity changes. It needs more practice to read, and species and season both shift the values. They find the two together more reliable than either alone.
Their time study gives about 52 minutes for one operator to scan an average tree with both instruments, and about 37 minutes per tree for two operators working through a queue. The cost of a scan comes to roughly EUR 25 to 30.
Instruments named in the studies
Two tomographs were compared on red pines that supported a ropes course in Michigan, in a U.S. Forest Service report. Its PiCUS tomograph with 12 sensors gave higher resolution of decay areas than its Fakopp 2D unit with eight sensors, and the two correlated well on approximate location.
| Instrument | What it measures | Sensors or points | Study |
|---|---|---|---|
| PiCUS Sonic Tomograph (Argus Electronic) | Sound transit time | 8 to 12 | Gilbert and Smiley 2004 |
| PiCUS 3 Sonic Tomograph | Sound transit time | 12 | Baláš, Gallo and Kuneš 2020 |
| PiCUS TreeTronic electrical resistance tomograph | Electrical resistivity | 24 | Baláš, Gallo and Kuneš 2020 |
| Fakopp 2D Acoustic Tomograph | Sound transit time | 8 | Allison, Wang and Ross, Forest Service |
Sources
- Gilbert and Smiley, Picus Sonic Tomography for the Quantification of Decay in White Oak and Hickory, Arboriculture & Urban Forestry 30(5), 2004, read October 15, 2026
- Gilbert and others, Use of sonic tomography to detect and quantify wood decay in living trees, Applications in Plant Sciences 4(12), December 2016, abstract, read October 15, 2026
- Baláš, Gallo and Kuneš, Work sampling and work process optimization in sonic and electrical resistance tree tomography, Journal of Forest Science 66(1), 2020, read October 15, 2026
- Allison, Wang and Ross, Visual and Nondestructive Evaluation of Red Pines Supporting a Ropes Course, USDA Forest Service, read October 15, 2026