Using GPR to estimate tree root biomass

USDA Forest Service (FS) researchers are improving the use of ground-penetrating radar (GPR) to study tree roots nondestructively. They are refining GPR’s processing capabilities by comparing results with those of more invasive methods.

GPR is an electromagnetic imaging technique that can be used to detect buried objects or hidden structures. GPR has been used for geological research, archaeology, forensics, and for assessing the integrity of roads and bridges. FS researchers soon recognized the potential for using the technology in forest-based research.

Measuring the belowground growth of trees is essential to understanding forest productivity and carbon allocation. Estimating the biomass of tree roots traditionally involves using soil cores, pits, and trenches–digging up roots, then sieving, washing, drying, and weighing them. These methods are destructive, labor-intensive, and not very useful for measuring the lateral extent of a root system.

In the September/October issue of the Soil Science Society of America Journal (SSSAJ), researchers from the FS Southern Research Station (SRS) unit in Research Triangle Park, NC present the results of a study that assesses ground penetrating radar (GPR) as a fast, noninvasive method to improve estimates of root biomass.

“Knowing both the volume and extent of root systems is important in the carbon sequestration studies we do,” says Kurt Johnsen, director of the SRS Biological Foundations of Southern Forest Productivity and Sustainability unit, and co-author of the article with John Butnor and Lance Kress. “Many of the forests in the Southeast grow on land where the soil carbon has been depleted by former farming practices. In these forests, tree roots are the most dynamic pool for carbon accumulation below the ground.”

For the carbon flux experiments that Johnsen and fellow researchers conduct at the Free Air Carbon Dioxide Enrichment (FACE) sites in the Duke Forest, they use a sophisticated dynamic gas sampling system to measure the effects of elevated levels of carbon dioxide on living trees. Although they can detect variability above the ground without harming the trees, it is almost impossible to know what is going on below ground. “We need a way to measure how the root system is responding that does not involve destroying it,” said Johnsen.

John Butnor, SRS plant physiologist and lead author of the SSSAJ article, has been experimenting with ways to make GPR more accurate by improving the quality of the data through advanced processing techniques, and by calibrating GPR estimates with those from soil cores.

“There are a variety of factors that can affect the resolution of radar profiles of roots. Soil composition can cause background noise that interferes with resolution and alters the results,” says Butnor. “For this study, we wanted to look at the full potential of GPR, so we chose a site with soil composition amenable to radar investigations–one with electrically resistive soil of high sand content.”

In collaboration with Lisa Samuelson (Auburn University), Butnor and the other researchers used a previously established International Paper study site in Georgia, setting out sample points on plots of loblolly pine that had been fertilized or irrigated or both. For GPR sampling, they passed the radar antenna across in one direction, then the other, electronically marking sampling points on the radar profile. When they finished GPR sampling, the researchers collected soil cores at the sampling points, weighing the washed and dried roots to determine total live biomass.

Butnor found that adding advanced digital processing techniques greatly improved the ability of GPR to accurately estimate root biomass. He was also able to correct for the distorting effects he found in the fertilized plots.

“By closely matching the footprint of the radar antenna to the location of the soil core, we were able to improve root biomass estimation significantly over our previous studies,” says Butnor. “The ability to correlate radar data to actual root biomass gives greater confidence in the technique and allows us to continue to make improvements.”

The researchers concluded that, in the right conditions, GPR can be used to rapidly estimate root biomass, dramatically reducing the number of soil cores that are usually needed and providing a much clearer picture of the lateral root system as it spreads out beneath the ground.

“We have shown that GPR works very accurately on well-drained soils,” says Johnsen. “In a four-hour period, we can collect as much data using GPR as collected from thousands of core samples. More recently we have used GPR on flatwood sites in Florida and on heavy organic matter sites in Canada. We believe that GPR will become a standard tool in forest research, and will someday allow us to do rapid, nondestructive root assessment across many soil types. ”