GPR is an effective tool for mapping buried tree roots and is a useful tool for city planners, arborists, and climate change scientists.
Many people are interested in imaging tree roots. A common application we hear is GPR being used by arborists to try and save an urban tree whose roots are interfering with city infrastructure such as underground pipes and cables. Cities are interested in saving trees where possible and knowing the critical root structure assists with that decision (Figure 1). More recently, climate change researchers are interested in tree roots as one component of underground biomass. They need a better understanding the structure of root systems and to measure root volumes to estimate how much carbon is sequestered in trees.
GPR is one of the few technologies that can detect in situ tree roots with high resolution. Surprisingly, the properties of tree roots are not that different from other objects that GPR is typically used to detect – non-metallic utilities. Specifically, detecting a tree root is very similar to detecting a water-filled plastic pipe. In both cases, the material that provides the contrast that GPR needs to reflect signal is water (Figure 2).
Depending on how dry the soil is, the reflectivity of a tree root can be close to 50%, which is very high for GPR. This is why tree roots can appear as very strong reflectors in GPR cross-sections (Figure 3).
When GPR data are collected in a grid or a pseudo-grid over an area, the data can be processed into map images called depth slices. Figure 4 is a depth slice of 500 MHz GPR data showing roots from mature trees at a park. The objective of this GPR survey was not to image tree roots; that was just a fortunate happenstance. The intended targets were graves located deeper than the tree roots. The data from this survey was featured in our July 2022 newsletter here: https://www.sensoft.ca/blog/historic-graves-located-in-a-modern-city-park/
The same grid and pseudo-grid data can also be displayed in 3 dimensions to show roots that vary in depth (Figure 5).
The key to generating clear and interpretable depth slices of tree roots is a very tight line spacing. People often make the mistake of collecting GPR data with a coarse line spacing. This leaves gaps in the data and, since tree roots don’t typically grow in straight lines like utilities, makes it difficult to interpolate data to fill the gaps and show the actual root paths. So, it is better to spend more time in the field collecting more data, increasing the data density, than spending hours in the office trying to interpret low-density GPR data.
Ideally, GPR lines should be spaced no greater than the length of the GPR antenna. For example, a 1000 MHz center frequency antenna is 7 cm long so, a line spacing of 10 cm provides excellent detail. For 500 MHz center frequency GPRs, a 15 to 20 cm line spacing is ideal. If you are not certain of the length of the antenna, use the width of the GPR sensor, the dimension perpendicular to the survey line direction.
For example, the tree root data shown in Figures 4 & 5, was collected with a 500 MHz GPR system and a line spacing of 25 cm. The images would have greatly benefitted from a tighter line spacing of 15 cm.
If you have any tree root data that you’d like to share with us, especially data that was subsequently ground-truthed by excavating the roots, please contact sensoft_training@spx.com
