Sonic and Electric Impedance Tomography
Sonic and Electric Impedance Tomography is used to produce color-coded images (tomograms) of the interior of tree trunks and large branches – both tree wood density (sonic) and wood hydration (electrical impedance). The tomograms produced are similar to an MRI images used for human medical imaging. These images can be useful in helping to assess the structural integrity of the scanned portion of the tree. The tomograms above show some sonic scans that are placed to the right of the photographed tree trunk cross-sections that they have imaged. We use the PiCUS™ sonic and TreeTronic™ electric impedance instruments and analytical software manufactured by Argus Electronic of Germany for our tomographic work.
Since 2010 our firm is one of the first to provide sonic tomography of trees in the Western USA and the first to offer electric resistance tomography.
Sonic Tomography tracks the speed of sound waves through the trunk or large branch of a tree. The tomogram is an image of the scanned area, which is more or less a flat plane cross-section through the tree or branch. The tomogram is calculated by measuring the speed at which sound moves between measuring points around the trunk or branch of a tree. To measure the speed nails are placed around the circumference of the tree. Next the shape of the tree is measured using a caliper so that the distance between all measuring points is known. Next sensors, that "listen" for the sound of a hammer tap are attached to each nail. Each measuring point is tapped with an electronic hammer. The instrument measures the time for the sound to travel from the measuring point being tapped to the all receiving sensors. After all sensors have been tapped and the time for the wave to travel across the tree is measured a tomogram is calculated producing a color coded image representing the velocity of the sound wave thru the tree. By interpreting this graph we can determine what the internal structure of the tree is. Sonic tomography detects differences in the ability of wood to transmit sound waves. It cannot identify the specific reason for such differences, however. This means that the instrument is not able to identify the exact type of damage -- the damage may be decay, hollowness, cracks or something else. Please note that using current sonic tomography it is very difficult to image roots.
The photo at right shows sonic sensors being placed around a 53-inch diameter California bay laurel tree, Umbellularia californica, in preparation for a sonic scan.
The photos on this page show sonic scanning of a black acacia, Acacia melanoxylon. The upper left photo shows the lower 40 feet of the tree, which leans over a lawn and parking lot area. Tim is placing the sensor belt around the trunk of the tree at about 4 feet above the ground. The upper right photo shows the scanners placed around the lower trunk. Arrows point to a large crease or seam that arises from an area of obvious basal decay (lowest arrow). The lower left photo shows a Ganoderma applanatum conk near a sensor. A conk is the fruiting body of a wood decay fungus (like a mushroom). Conks usually don’t appear until after there has been extensive interior decay within a trunk or branch.
These scans and photos are from the black acacia on the previous page. Clockwise from upper left: the sonic scan image, the electric impedance scan image and finally a photo of the cut-out cross-section of the trunk where the scans took place.
In the sonic scan the densest wood is indicated by a black with decreasing density as brown then green, violet and blue. The lighter shades are less dense than darker shades of the same color.
In the electric impedance scan the highest water content is shown by dark blue, then light blue, then green, yellow, red and brown. Each tree species had its own characteristic “normal” electric pattern.
Electric impedance Tomography (EIT) uses current/voltage to examine a tree. The resulting measurements are displayed in a two-dimensional image showing the apparent electrical impedance of the wood, called an electrical impedance tomogram. The tomogram is an average of the electric resistance/conductivity within a column of roughly half of the height of the diameter of the trunk or branch above and below the line of sensors. EIT measures the chemical properties of the wood, such as water content, ion concentration, cell structure, etc., which change according to the status of the wood. This test can detect such changes by measuring how the heterogeneous wood “bends” the electrical field. In several research studies this type of tomography has proven very useful in detecting early disease and decay in standing trees.
The electric impedance tomogram is somewhat species-unique in that there is a typical pattern of electrical conductivity and impedance that is characteristic for a particular tree species. It is important to know what that pattern is when interpreting an EIT so that an abnormal pattern can be identified for that species. Since there is no track record of use in California, we are just beginning to collect data on normal EIT’s for various tree species. When we come upon a tree species that we have not tested yet, it is preferable to test a healthy tree of the same species nearby, or (second-best situation) a section of the trunk or branch of the subject tree that appears defect-free, in order to establish a normal baseline EIT for that tree species.
Combining Sonic and Electric Impedance Scans: Combining the sonic and electric impedance tomograms provides more information about the type of defects in a tree and their location, which is why we generally perform both types of scans on the trees that we test. This is relatively simple because the same tree geometry that has been measured for the sonic scan can also be used for the electric scan, although additional more sensor nails must be added for an EIT. Evaluating the two measurements together helps overcome some of the limits of sonic tomography, such as “acoustic shadows” caused by cracks. Of the few tree tomography instrument manufacturers that exist, Argus Electronic of Germany is the only supplier of a sonic as well as an electric impedance tomography instrument, which is why we purchased our equipment from this company.
Accurate tree geometry is important in tree tomography! The cross-section of trees is not always a perfect circle! Therefore we make many measurements between sensor locations with a large electronic caliper, in order to establish a very accurate cross-sectional geometry of the measuring level of the tree we are scanning. The caliper sends the measurements directly to our computer that we are using in the field, so that there are no transfer errors. Accurate geometry translates to better tomograms!
3-D tomograms are also possible, by conducting multiple scans at different levels on the tree. The areas between the measuring levels are interpolated. The image can be rotated so that all sides of the trunk or branch can be seen in 3D.
For a sample report click here. Sample tomography report.
Link to presentation on Tree Tomography presented at the California Tree Failure Report Program Annual Conference Jan 9, 2014
The IML Resistograph™ measures the penetration resistance of a specialized drill bit, producing a visual graph of wood density. The graphic printout (the drill trace) is often helpful in estimating and comprehending the interior status of the tree when taken in context with other tree and site factors. The Resistograph can also be used in tandem with other testing methods such as tomography. The graphs produced can be quite species specific however, and one of the problems with the Resistograph is that there is not enough good data on many tree species to be able to interpret the graphs with confidence.
The photo above shows an F-400-S Resistograph with a 15-inch drilling depth.
The upper photo shows an IML 400-S Resistograph with a 15-inch depth drilling capacity. The lower photo shows Tim using the Resistograph in the field on a large coast live oak, Quercus agrifolia. Drills are being performed around the tree at the level of the yellow tape. Drill locations are tagged with number flags, such as drill #3 shown at the right in the photo. The base of the trunk of this tree was buried by about 18 inches of fill soil, and is suspected to be decayed because of this. Drill test results are shown below.
The Resistograph trace graph above shows extensive decay and a flat line (cavity from about 6.5 inches (arrow) to the end of the drill at 15 inches. The trace below, taken on the same tree in a healthy area shows a normal trace for healthy wood of this particular tree species, with a slight decline (possibly early decay?) at about 11 inches (arrow) to the end.
Different tree species will produce different drill traces, so it is important to understand what is “normal” for a particular species in order to recognize what is abnormal for that same species. Older versus younger trees of the same species will often produce different traces because younger wood is usually less dense than older wood. Younger trees have more young wood near the outside of their trunk versus the interior of the trunk, compared to older trees. There are other idiosyncrasies that influence differences in traces between trees.
For all of the above testing instruments and methods it is important to note that each test “sees” only a very small portion of the entire tree (a drill profile for the Resistograph, a flat plane cross-section of the tree for sonic tomography, and the average of a column above and below the sensor line for electric impedance tomography) – so it is possible that defects that are important to tree structural stability can be missed or misinterpreted in the scan images. Although multiple tests at different levels and locations can be performed, the entire tree cannot be drilled or scanned! Therefore it is not possible to certify that a tree is “safe” or “unsafe” by using the any of the above diagnostic techniques or any other diagnostic techniques, for that matter. The use and interpretation of diagnostic techniques such as Resistograph drilling and tomography is both an art and a science that is still evolving and probably always will be! The tree owner must accept all risk associated with the use of the diagnostic equipment described herein, the interpretation of test results and recommendations based upon those interpretations and results.
Tomographic scan images of trees cannot tell us what has caused a deviation in a tomogram – all it can tell is the variation between the travel speed of sound (sonic tomography) or electric conductivity and resistance (electric impedance tomography) within the tested area. Use of tomography may provide good information that is helpful in making a decision about a tree; and other times it may not. It possible to misinterpret a tomogram, and sometimes even if a tomogram is interpreted correctly false positives or negatives may be generated). As with all other diagnostic instruments, the information provided by tomography cannot be guaranteed and must always be taken in context with other factors (such as tree species, outward signs or symptoms of internal defects, the tree owner’s tolerance of risk, financial considerations, etc.) With electric impedance tomography our ability to detect abnormal patterns is limited at this time unless we have sufficient background normal EIT data for the particular tree species that we are scanning. We are currently in the process of collecting and cataloging this data!
Tomography performed at less than 0.5 to 1 times the trunk diameter (e.g. the tree’s root collar) may have reduced accuracy due to gaps and included bark between roots, soil interference and limitations of the technology. This is especially true for electric resistance tomography, but the accuracy of sonic tomography can also be reduced in this zone. Sometimes we do perform tomography of the root collar area however, because it is of interest and evaluation of the electric and sonic scans together often provides some useful information. It is important to note however, that interpretation of tomograms in this area is more difficult and prone to error, and the client must accept this condition.
A few “quirks” of tomography in trees:
The interpretation of tomogram is difficult and requires sound working knowledge of trees and their diseases as well as the working principles of the measuring instruments. A tree should not be condemned solely based upon a tomogram reading! It is important to analyze all possible reasons for defects shown in tomograms, and different investigative methods should be included as necessary in order to develop an intelligent conclusion and recommendations for a particular tree. Tomogram interpretation requires a great deal of training and experience and should not be attempted by unqualified persons. A tomogram in the hands of the wrong person can be a dangerous thing!
Like tomography, information obtained from the use of the Resistograph drill on trees will sometimes provide information that is helpful in making a decision about a tree; and other times it will not. It is possible to misinterpret a Resistograph drill trace. False positives or negatives may also be generated due to idiosyncrasies in both the wood and the operation and performance of the instrument. As with all other diagnostic instruments, the information provided by the Resistograph cannot be guaranteed and must always be taken in context with other tree-related factors and the site.
The Picus sonic and electric impedance tomograph (as well as a competing sonic-only tomograph device developed by another manufacturer) were developed in Germany and have been used most extensively in Europe and secondarily in the Northeast United States. Since there is no good track with the use of this machine on both native and non-native tree species in California, we have started to collect this data. We are partnering with Lothar Goeke of Argus Electronics, Germany, to create a database of sonic and electric resistance images and numeric data such as sonic and electric amperage/voltage ranges that can be used for reference and comparison of different tree species around the world. We have also been assisted in our work by Dr. Larry Costello, U. C. Horticultural Adviser, emeritus, to whom we are very grateful.
For the first six months after we acquired the instrument, in order to best learn how the instrument works and how to interpret scan images, we sought out trees in our area that were scheduled to be removed. Prior to removal we performed sonic and electric impedance scans on these trees and also compared the performance of tomography with Resistograph drill testing on some of the trees. As the trees were removed we cut apart the scanned and drilled sections of the test trees in order to see how well the actual interior matches the tomographic scans and drill trace graphs. This is a “dissection test”. We tested several different tree species in this manner, but the most numerous species in this pilot study was coast live oak, Quercus agrifolia and we have a good amount of data and experience for this particular tree species at this time. We are grateful to Jess Running of the Davey Tree Service Company and Jim Lewis of Tree Team Tree Service for their assistance in locating and obtaining permission to scan these trees, and also in their help in cutting apart the trees for dissection and study.
Palms too! We are also partnering with Don Hodel, U. C. Horticultural Adviser for Los Angeles County, on utilizing tomography on palm trees, which have a very different physiology than “regular trees” such as hardwoods (e.g. oaks) or softwoods (e.g. conifers like pines and redwoods). Tomography may be useful for detecting serious problems such as Sudden Crown Drop of Canary Island palm, Phoenix canariensis, before a failure occurs.
Please see the Technical information on PiCUS tomography of my Web Site. I update this from time to time, so check back frequently!
How much time do tomography scans take to perform?
These are not quick, easy, cheap out-of-the box tests! If the diameter of the scanned area of the tree is 24 to 36 inches expect 1.5 to 2.5 hours per combination sonic/electric impedance scan per tree. This assumes a typical trunk scan 6 feet or less above the ground with easy access to the tree for a vehicle and the instrumentation. If access is more difficult, then additional time will be necessary. For larger trees (e.g. scan diameters more than 36 to approximately 60 inches) expect 3 to 4 hours per scan per tree. Large trees with scans at multiple levels may require an entire day or even multiple days. Scaffolding or other support devices are required for scans 7 ft or more above the ground. It is recommended that the testing be started in the morning so as to allow as much time as possible to complete the work in one site visit.
The time taken to run the scans themselves is much less than the time that is necessary to set up the process beforehand. Pre-scan work includes measuring the geometry of the area to be scanned, place the sensors on the tree and enter the appropriate information into the computer program so that a scan image can be generated. We measure the geometry of the tree with an electronic caliper that is extremely accurate and generates a very realistic shape of the area to be scanned. The more accurate the initial geometry you start out with, the more accurate the scan! Sonic scans are run first and these have the longest set-up time. The electric impedance scan is run after the sonic scan, and it can utilize the same tree geometry measurements and most of the other information that was either input or generated by the sonic scan. Additional sensor nails are inserted for the electric scan so that there will be exactly twice as many electric as sonic sensors.
In addition to performing Resistograph drills, tomographic scans and other tests, we will also visually examine the tree and assess the site. Other factors besides test results need to be considered order to most intelligently decide what to do with a particular tree! Oftentimes we have already done this pre-testing “homework” on a previous site visit and evaluation of the tree – before we recommended that this additional testing be performed.
How much time do Resistograph drill tests take to perform?
It takes only a few minutes to perform a drill test, but more time should be spent in selecting and documenting the test location and results through field notes, sketches and photographs. Depending upon the number of drill tests performed and the goal (for example trying to estimate the percentage of sound wood all around the perimeter of a tree) this can take up to several hours per tree. When the Resistograph is used in combination with prior tomographic scans of the tree, then drilling time and analysis will take less time because there is already an accurate scale drawing of the perimeter of the area from the tomography trunk diameter measurements. Fewer drill tests are necessary because information from the tomograms is used to determine where to place the drill tests in order to obtain the most useful information.
In addition to the actual time in the field performing the tests, additional time in the office is necessary to review the images and the raw data behind them, and also to produce a written report with our findings, opinions and recommendations.